H |

C

O

2 -i 2 _ .

IfUllSNI NVINOSHJ.il/MS S3iaVaan LIBRARIES SMITHSONIAN INSTITUT

n , Z r* 2 r~

ro ° 5 /sds^s. - in /jCsSSl

co _ co £ co

5 R ARIES SMITHSONIAN INSTITUTION NOIirUllSNI NVIN0SH1MS S3IUVH1

z v co 2 co

—i X CO

o

uwji!/ 2 >

> 2 ' \^*' > ' 2 Xioosv^z > v

IfUllSNI NV1N0SH1IWS^$3 I d Vd a n\| B RAR I ES^SMITHSONIAN^INSTITUT ” </> ^ co =

XJF O 5 «•' * X 5 N%£d

IRARIES SMITHSONIAN INSTITUTION NOIIIUIISNI "WlNOSHillAIS ^S3 I a va z r~

03 TO >

70

— vc;

ro

ixniuSNi^NviNosHiiws^ss i avaa n~Li b r ar i es "smithsonian “institut

s > -vr 5 » >

U5 Z </> ♦ Z y,

5RARIES SMITHSONIAN INSTITUTION N0IJ.rUIJ.SNI NVINOSHilWS S3iaVH

. ^ CO T* ^ ^ ... CO

A^O'V/X W (n <-U £ ^C\ N UJ

5 ® o

iniUSNI^NVINOSHillNS SaiavaaiT^UBRARI ES^SMITHSONIAN^INSTITUT

2 r- 2

m ' W ~ x^ast^ m xi\osv^>2 ^ ro

CO £ CO £ co

IRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHJLIWS S3 I H V d S * 2 . z: * co 2 co

> 2 X^osv^x >

IfUllSNI NVINOSHIIINS^SB I dVH a nZLI B RAR I ES^SMITHSONIAN^INSTITUT

nillSNI^NVINOSHillNS^SB I d V d a 1 1 ^ L I B R A R I ES^ SMITHSONIAN^INSTITUTIO r~ z “ ~~

o

m v^v - m

co - co ± C r>

R ARIES SMITHSONIAN INSTITUTION NOlinilJLSN! NVINOSHillAiS S3I9V99

CO Z *■ CO Z CO

2 2 s

<j\

I t z /V/' t ^ubjww =.»

LnilJLSN! NVINOSHIMS^SB I dVd 9 n\l BRAR I ES^SMITHSONIAN^INSTITUTIC

co _ ” CO —

RARIES SMITHSONIAN INSTITUTION NOlifUllSNI NVINOSH1IWS S3ldVda

~ 2 r* . z

m

CO ™ CO

millSNI NVINOSHillAiS S3ldVdail LIBRARIES SMITHSONIAN INSTITUTE

:,< 5 /// /w**- ^

I 9 4§k*X £ (?£

1 ^

> W" 2 >

RAR I ESC°SM1THS0NIAN INSTITUTION NOliniliSNI NVINOSHilWS^Sa I d V d a co =; co — co

Q __ N^UIJL^ Q

LniliSNI^NVINOSHillNS S3 I dVd 9 ll^LI B RAR I ES^ SMITHSON I AN™* INSTITUTE r~ , Z r* Z r-

m ' vs^ 5^ m

co co

RARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHillNS S3ldVd9

CO z * co z CO

2 X^v'TO/X < A 2 <

S /®p"*Sv I ^

fyF'/ffly'' x tafe- o

> > '■ ^ ^ inillSNI NVINOSHimS^SB I dVd a ll^LI BRAR I ES^SMITHSONIAN INSTITUTK

PSYCHE

A Journal of Entomology

Volume XXXII 1925

Edited by Charles T. Brues

Published by the Cambridge Entomological Club, Bussey Institution, Forest Hills, Boston 30, Mass., U. S. A.

Printed by The St. Albans Messenger Company St. Albans, Vermont.

PSYCHE

A JOURNAL OF ENTOMOLOGY

Established in 1874

VOL. XXXII FEBRUARY, 1925 No. 1

TABLE OF CONTENTS.

TABLE OF CONTENTS.

A Note on the Moulting of the Tarantula. Eurypelma hentzii

Phil Rau 1

New Nemestrinidse (Diptera) from Rhodesia and New Guinea

J. Bequaert . 4

Some Species of the Genus Leucospis. C. T. Brues 23

The Efficiency of Birds in Destroying Over- Wintering Larvae

of the European Corn Borer in New England. G. W. Barber , ... 30

A New Species of the Genus Gaurax. C. W. Johnson 47

New Neotropical Thysanoptera Collected by C. B. Williams.

J. D . Hood 48

Some Hitherto Undescribed Habits of Meskea dyspteraria Grote.

R. L. Schwartz 70

CAMBRIDGE ENTOMOLOGICAL CLUB

	OFFICERS FOR 1924
President ,	C. T. Brues
Vice-President	R. H. Howe, Jr.
Secretary	J. H. Emerton
Treasurer	F. H. Walker
Executive Committee	. A. P. Morse, S. M. Dohanian
	S. W. Denton

EDITORIAL BOARD OF PSYCHE

EDITOR-IN-CHIEF

C. T. Brues, Harvard University

ASSOCIATE EDITORS

C. W. Johnson, Nathan Banks,

Boston Society of Natural History. Harvard University.

A. L. Melander, A. P. Morse,

Washington State College, Peabody Museum.

J. H. Emerton, J. G. Needham,

Boston, Mass. Cornell University.

W. M. Wheeler,

Harvard University.

PSYCHE is published bi-monthly, the issues appearing in February, April, June, August, October and December. Subscription price, per year, payable in advance: $ 2.00 to sub- scribers in the United States, Canada or Mexico; foreign postage, 15 cents extra. Single copies, 40 cents.

Cheques and remittances should be addressed to Treasurer, Cambridge Entomological Club, Bussey Institution, Forest Hills, Boston 30, Mass.

Orders for back volumes, missing numbers, notices of change of address, etc., should be sent to Cambridge Entomological Club, Bussey Institution, Forest Hills, Boston, 30, Mass.

IMPORTANT NOTICE TO CONTRIBUTORS.

Manuscripts intended for publication, books intended for review, and other editorial matter, should be addressed to Professor C. T. Brues, Bussey Institution, Forest Hills, Boston, 30 Mass.

Authors contributing articles over 8 printed pages in length will be required to bear a part of the extra expense, for additional pages. This expense will be that of typesetting only, which is about $2.00 per page. The actual cost of preparing cuts for all illustrations must be borne by contributors; the expense for full page plates from line drawings is approximately $5.00 each, and for full page half-tones, $7.50 each; smaller sizes in pro- portion.

AUTHOR’S SEPARATES.

Reprints of articles may be secured by authors, if they are ordered before, or at the time proofs are received for corrections. The cost of these will be furnished by the Editor on application.

Entered as second-class mail matter at the Post Office at Boston, Maas. Acceptance or mailing at special rate of postage provided in Section 1103, Act of October 3, 1917, authorised on June 29, 1918.

PSYCHE

VOL. XXXII. FEBRUARY 1925 No. 1

A NOTE ON THE MOULTING OF THE TARANTULA. EURYPELMA HENTZIP

By Phil Rau,

St. Louis, Missouri.

A specimen of this spider was brought to me from Texas by a friend on April 15, 1922. It lived in confinement almost a year, and fed upon various insects which were placed in the cage, such as grasshoppers, Dissosteira Carolina, cabbage butterflies, larvae of the pipe mud-wasp, Trypoxylon politum, larvae of the green June-beetle and unidentified small moths. It refused, however, to eat adult May-beetles, Lachnosterna sp., dung beetles, Canthon Iceiis, centipede, Scutigera forceps, bugs belong- ing to the family Pentatomidae, and male wasps, Polistes pallipes. It is also possible for this species to go for long periods entirely without food. At one time when I was out of the city, a star- vation period of two weeks did not seem to harm it.

On one occasion I caught it in the act of eating a fat larva of the June-bug. The spider stood high up on its legs while under its jaws it held the large mass of meat which shortly before had been the larva. Upon repeated proddings, the spider walked slowly away carrying the morsel in its mouth. Finally under provocation the spider let go and then I saw that the food had been reduced to a mushy mass. So thoroughly was it masticated that only by a small portion of the skin was I able to learn its identity. That the tarantula actually chews its prey was demonstrated in the case of the larva of the mud wasp also, but whether the spider actually eats these food masses or only sucks the juices, I do not know.

This spider lived an uneventful life, with the exception of its maneuvers of moulting. This process took place on August

identified by J. H. Emerton.

2

Psyche

[February

13. For two weeks previous to this date, the spider had refused all food. At eight o’clock that morning it seemed fairly lively. At 2:30 p. m. when I again tried to tempt it with food, I found it lying prostrate on the floor of the cage, with the legs stretched out flat. When I attempted to place it in preserving fluid, I found signs of life, and after watching over it for fifteen minutes I found that the spider was actually in the throes of moulting. First the carapace cracked at the sides and along the front, and fell back on the abdomen and lay there inverted. Then by bodily contortions the skin was slowly slipped off the abdomen, or rather the abdomen slowly emerged from the old skin. At this stage all of the legs, as well as the mouth-parts, were still in the old skin, and it was puzzling to me to guess just how they would be shed. Up to this time the legs had been spread in a very natural, free position, but before the abdomen had com- pletely left the old skin I noticed that the spider was gradually raising itself up, up and up, gradually pulling itself out of the old legs, palpus and chelicera coverings. It was a beautiful process to observe. All eight legs were lifted simultaneously, and soon the spider lay helpless on its side, entirely free from the old skin, all limp and clean, and beautiful as new velvet. The entire process of moulting, from the time that the skin began to crack at the shield to the complete extrication, took twenty minutes. Fig. 1 shows the shed skin intact, just as the spider left it. C — carapace, DC — dorsal covering, v. c. — ^ventral covering, F — one of the fangs removed from the main portion for photo- graphing.

The spider after moulting was very pleasing to the eye. While the carapace had the same grayish-white color, the ab- domen had a beautiful covering of silvery brown hair. Some change had occurred in the color of the legs and chelicerae, for while they had been brown, after moulting they were of a slate gray color, and remained thus without change until the death of the spider, five months later. However, three weeks after moulting it was noticeable that some of the hairs on the front legs had changed to brown. I suppose that the slate color is not a permanent character, but one that changes with age.

Another item of interest was the behavior of the dorsal

1925]

A Note on the Moulting of the Tarantula

3

groove on the carapace at that time. Under normal conditions, this seems tightly closed, but during moulting it pulsated slowly and constantly, in a way suggestive of air being pumped in. This groove is the point of attachment of the thoracic muscles.

Fig. 1. Moulting Tarantula, Eurypelma hentzii.

The animal was probably adult after this moult, and the hazard of moulting had been surmounted.

One other item of interest should be recorded in connection with this narrative. Into this large glass box that served as its house I at one time placed a spider, Latrodectus mactans. This creature made a straggly web in one corner, and the tarantula often used it to climb to the top, a distance of fifteen inches. It picked its way carefully and slowly among the strands of web, carefully placing a foot here and lifting one there, as with great dexterity, it lifted its ponderous body, thread by thread, among the apparently insufficient threads.

4

Psyche

[February

NEW NEMESTRINIDtE (DIPTERA) FROM RHODESIA AND NEW GUINEA

By J. Bequaert.

Department of Tropical Medicine, Harvard University Medical

School.

The curious and apparently archaic family Nemestrinidae is rather abundantly represented in South Africa, but very few species are known north of the Orange and Limpopo Rivers. It is, therefore, of much interest to record three new forms, of the genera Prosceca and Stenobasipteron, which have been re- cently discovered in Southern Rhodesia. I wish to thank Dr. G. Arnold, Curator of the Rhodesia Museum, Bulawayo, for the opportunity of studying these insects.

. On this occasion I shall also describe a new species of A yct°rimyia, from New Guinea, entrusted to me some time ago by the Paris Museum.

Prosoeca rhodesiensis sp. nov.

Type female from Matopos, Southern Rhodesia, April 17, 1923 (R. Stevenson Coll.); allotype male from Mt. Bambata, Matopos, Southern Rhodesia, March 23, 1924 (without collector.) Both in the collection of the Rhodesia Museum, Bulawayo.

A robust, black species, covered with dull grey tomentum; vertex and dorsum of thorax with short black hair; pilosity otherwise greyish white, very long and dense on the under side; a dorsal row of brownish black, dull spots on the middle of ab- domen; legs dark clove brown. Wings of normal shape in the male, with all longitudinal veins turned up at apex; brownish along costa and gradually fading into the hyaline hind margin.

Female : Integument black, faintly clove brown at extreme lower apex of face. Antennae, palpi, and proboscis black; the proboscis faintly clove brown toward the base. Legs very dark clove brown, the tarsi and claws almost black.

Body short pilose above, densely hairy on the ventral side. Vertex with erect, black hairs as far as the anterior ocellus; the remainder of the head with white pile, which is extremely short

1925] New N emestrinidce from Rhodesia and New Guinea

5

on the front, longer on the face and posterior orbits, and very long and dense on the cheeks. Dorsum of thorax with moder- ately long and rather sparse., erect, black pile; scutellum with similar, but somewhat longer, black hair, except behind its posterior margin, where the pilosity is greyish white; sides and ventral face of thorax densely covered with long, soft, greyish white hairs, which extend as a distinct white stripe above the base of the wing. Abdomen dorsally with sparse and short, erect, black pile; at the base and along the hind margins of the segments there is a mixture of greyish white hairs; ventrally the pilosity is longer, den'ser, greyish white, and generally appressed. Coxae and femora with long, greyish white hairs; the pilosity of the tibiae and tarsi extremely short, black; the longer setae at the tip of the tibiae also black. Except where the pilosity is very long and dense, the integument is covered with a dull, ashy grey bloom; on this, one may see, in the proper light, two wide, longitudinal stripes of blackish pruinescence in the anterior half of the thoracic dorsum, on each side of, and close to, the middle line. Brownish black pruinescence also forms a row of median, rounded, dull spotb on the second, third, and fourth abdominal tergites; each spot being located close to the anterior margin. In the female I have seen, these spots are quite well marked.

Head large, flattened, much broader than the thorax; semi-elliptical in profile; kidney-shaped and nearly one and one-half times as wide as high when seen in front. Front rather narrow, widest at the insertion of the antennae, where it measures about half the width of the eye; the inner orbits converge dis- tinctly toward the aiiterior ocellus, wheire the front is only half as wide as at the antennae. Vertex nearly parallel-sided. Ocellar protuberance elongate and low, but slightly separated from the inner orbits, with a transverse, saddle-shaped depression in the middle; ocelli placed in an isosceles triangle, the posterior ocelli being only about half as far from each other as from the an- terior ocellus. Eyes bare. Antennae short, small, placed on the sides of the face, close to the inner orbits; basal segment subcylin- drical, slightly longer than wide, broadly truncate and some- what emarginate at apex; second segment nearly as long as wide, about two-thirds the length of the first, squarely truncate

6

Psyche

[February

at apex, with rounded edges; third segment flattened, pear- shaped, but little shorter than the first and second segments together, twice as long as wide, broadest in its basal half and thence gradually narrowed to the truncate and slightly sinuate apex. Style longer than the whole antenna, very sharply three- jointed; the two basal divisions thick, of about equal length, together about two-thirds the length of the third antennal seg- ment. Front slightly convex between ocelli and antennae. Face moderately swollen as a whole, gradually slanting from between the antennae to the oral margin, without grooves. The lower part of the head distinctly excavated between the cheeks. Proboscis of medium length, reaching about to the hind margin of the scutellum if supposed folded beneath the thorax; rather thick, especially in its basal half; directed downward, with a slight posterior slant. Palpi short and thick, three-jointed; the second segment much the longest; the apical segment bluntly truncate. Body quite broad and heavy. Thorax distinctly broader than thick; dorsum slightly wider than long; transverse suture quite deep on the sides over one-third of the width of the dorsum, continued as a shallow, oblique depression to near the scutellum. Scutellum large, semi-elliptical, cushion-shaped, its posterior margin separated from the disk by an impressed line. Abdomen broad and flat; the four basal segments together shorter than wide; the succeeding apical segments much nar- rower, decreasing in width, partly retractile within one another to form a telescope-shaped ovipositor; the last segment ends in two short, slender, straight, bluntly pointed lamellae which are wider in their basal half (in profile). Legs stout and long; the tarsi especially thick; the hind basitarsus but little narrower than the hind tibia.

Wings rather long and narrow, much longer than the body, over three times as long as wide. Costal cell and extreme base of wing, as far as the branching of the fourth and fifth longitudi- nal veins, infuscate, with a brownish yellow tinge; then fading into greyish in the first basal and subcostal cells and also along the costa to the tip of the wing; the remainder of the wing, in- cluding the alula, nearly hyaline. Veins dark clove brown or nearly black. Epaulet and basicosta clove brown; the epaulet

1925] New N emestrinidce from Rhodesia and New Guinea

7

with a tuft of long, appressed, white hairs. Venation of the usual Prosceca- type; all the longitudinal veins turned upward to end before the apex of the wing; no cross-veins between the terminal branches of the fourth vein nor between the second and the upper branch of the third; fourth posterior cell sessile.

Length not including ovipositor (to apex of tergite 4), 16.5 mm; greatest width of abdomen, 8.5 mm; length of proboscis, 8.5 mm; length of wing, 21 mm; width of wing, 6 mm.

Male. Very similar to the female in every respect. The abdomen is more clove-brown than black, but this is probably due to the fact that the greyish bloom is not as well preserved in the specimen in hand; the dull, black spots on the middle of the abdomen are present, though not quite as distinct as in the female. The vertex is just a trifle narrower at the anterior ocellus than in the female; but the ocellar triangle is still much longer than wide behind. The wings are slightly wider and a little more infuscated than in the female; but there is no pro- minent thickening of the costa beyond the middle and the passage to the nearly hyaline hind portion of the wing is quite gradual. The venation is as in the female.

Total length, 18.5 mm.; greatest width of abdomen, 9 mm.; length of proboscis, 9 mm. : length of wing, 23 mm.; width of wing, 6.7 mm.

This species is closely related to Prosceca beckeri Lichtwardt, of which it was at first thought to be but a variation. Owing to the kindness of Dr. H. Brauns, I was able to examine a male of P. beckeri from Montagu Pass, George, Cape Province. It differs from the Rhodesian male in several structural peculiarities which, however, could not be gathered from published accounts. Thus the wing of P. beckeri is, in the male, prominently widened beyond the middle, the costa being there considerably thickened (length of wing, 21 mm.; width of wing, 6.5 mm.); the wing being shaped somewhat like that of the males of Ommatius (Asilidse) and of Stenobasipteron (Nemestrinidse). Such a struc- ture of the wing is not found in the male of P. rhodesiensis. In addition, P. beckeri has the wing much darker in its anterior half; the vertex is quite broad, the ocellar tubercle shorter than wide, the two posterior ocelli being somewhat farther from each

8

Psyche

[February

other than from the anterior ocellus; the style is much shorter, being about as long as the whole antenna; there are also minor differences in the color of the pilositv, the hairs of the dorsum of the thorax being to a large extent greyish white, and there are two rows of blackish spots on the abdomen, instead of one row as in P. rhodesiensis.

Stenobasipteron arnoldi sp. nov.

Type female from Mt. Bambata, Matopos, Southern Rhodesia, March 23, 1924 (without collector); paratype female from the same locality and date. The type in the collection of the Rhodesia Museum; the paratype in my collection.

A slender, black species, covered with dull, cinereous tomen- tum; head and under side of abdomen paler; antennae and legs dirty straw yellow; pilosity sparse; longer and denser on the ventral side; proboscis considerably longer than the body. Wings very long and narrow, with the usual venation for the genus, very slightly smoky, more infuscate in the costal cell.

Female. Integument black on upper part of head and on dorsal face of thorax and abdomen. Face pale clove-brown. Sides of thorax blackish, with indistinct, yellowish brown blotches. Ventral face of abdomen pale dirty yellow. Antennae yellowish brown, the last division of the style black. Palpi clove-brown. Proboscis black. Legs entirely pale testaceous; claws black.

Pilosity sparse (probably but partly preserved in the two specimens seen). Vertex, front and face with very few, but long, erect, black hairs; cheeks and posterior orbits densely covered with long, greyish white pile. Dorsum of thorax and scutellum with sparse, but rather long, erect, black hairs; sides and ventral face, as also under side of scutellum, with more abundant and longer, somewhat yellowish white pilosity. Dorsally on the abdomen the hairs are mostly black, except at the extreme base; the anterior third of the second tergite has a sparse, erect, long, black pilosity ; the remainder of the dorsal side bears many scattered, extremely short, slanting, rather stiff, black hairs. Ventrally the abdomen is but poorly covered; there are a few,

1925] New N emestrinidce from Rhodesia and New Guinea

9

short, appressed, somewhat silvery white hairs, which are more abundant toward the sides. Coxae and femora with long, yellowish white pile; that of the tibiae and tarsi extremely short, white; the under side of the tarsi with more abundant, reddish brown pile. The integument of the entire body is covered with a dull, dark ashy grey bloom; on the front and face the pruines- cence has a slight yellowish tinge, and it is much paler, nearly white on the ventral side of the abdomen; there are no spots nor stripes on thorax or abdomen.

Head moderately flattened, much broader than the thorax; semi-elliptical seen from above; triangular in profile, due to the conically projecting face; kidney-shaped and nearly twice as wide as high in the middle, when seen in front. Front rather narrow, widest at the insertion of the antennae, where it measures about half the width of the eye; inner orbits distinctly converg- ing toward the anterior ocellus, where the front is but half as wide as at the antennae. Sides of the vertex slightly diverging behind. Ocellar protuberance short and low, about as wide as long, but slightly separated from the inner orbits; ocelli placed in an equilateral triangle; anterior ocellus over twice the size of each of the posterior ocelli, transversely elliptical, occupying more than half the width of the front. Eyes bare. Antennae (Fig. la) short, small, placed on the sides of the face, close to the

Fig. 1. Stenobasipteron. Right antenna drawn from the inner side: a, S. arnoldi; b , 5. difficile ; c, S. gracile.

10

Psyche

[February

inner orbits; basal segment cylindrical, nearly one and a half times as long as wide, squarely truncate at the apex; second segment a little over half the length of the first, about as long as wide, slightly broader at the apex which is broadly rounded off; third segment slightly flattened, short pear-shaped, but little longer than the first, and slightly over one and one-half times as long as wide, widest in its basal third and thence grad- ually narrowed to the straightly truncate apex. Style about twice the length of the whole antenna, sharply three-jointed; the two basal divisions together but little shorter than the second and third antennal segments; the second division some- what longer than the first. Front very feebly convex between anterior ocellus and antennae. Face much swollen, projecting anteriorly as a blunt cone, without grooves. The lower portion of the head is deeply and broadly excavated in the middle between the cheeks, the eyes continuing for about one-quarter their length below the oral margin. Proboscis very long and slender, reaching considerably beyond the tip of the abdomen when folded beneath the body, in which position it is in the type; while in the paratype it is directed downward with a slight anterior slant; labella thin and elongate. Palpi short and slender, distinctly three-jointed; the two apical segments much longer and of about equal length; the third truncate at apex. Body slender. Thorax about as broad as thick; dorsum dis- tinctly longer than wide; transverse suture deep on the sides over less than one-third the width of dorsum, continued back- ward to near the scutellum. Scutellum large, semi-elliptical, cushion-shaped; its posterior margin faintly separated from the disk by an impressed line. Abdomen flattened dorsallv, slightly wider than the thorax; the four basal segments together about as long as wide; the succeeding, apical segments much narrower, gradually decreasing in width, partly retractile as a telescope-shaped ovipositor. The last segment ends in two short, slender, straight, bluntly pointed lamellae. Legs long and thin; femora slightly swollen toward the base, more distinctly so on the front legs; tips of tibiae faintly thickened.

Wings very long and narrow, much longer than the body, over four times as long as wide. Costal margin nearly straight;.

1925] New Nemestrinidce from, Rhodesia and New Guinea

11

the posterior margin much constricted in its basal quarter, where a very narrow trace of the alula extends from the axillary excision to the base of the wing. Wings very faintly infuscate all over; more distinctly yellowish grey in the costal cell and at the extreme base. Veins dark clove-brown; epaulet and basicosta nearly black; the epaulet mostly covered with black pile. Venation of the usual type of the genus; fourth posterior cell with a long petiole at base; sixth longitudinal vein faintly undulate (more so than in S. gracile Lichtwardt); no “bulla” at base of second vein; axillary vein not developed beyond axillary incision.

Length not including ovipositor (to apex of tergite 4), 10.5 mm.; greatest width of abdomen, 4.5 mm.; length of proboscis, 14.5 mm.; length of wing, 14 mm.; width of wing, 3.4 mm. In the paratype these measurements are respectively 11 mm.; 5 mm.; 16mm.; 15.5mm.; 4mm.

This species is allied to Stenobasipteron gracile Lichtwardt, also of Southern Rhodesia. From published accounts alone, it would have been difficult to point out the differences. For- tunately, I was able to compare specimens of the two species. The proboscis is decidedly longer in S. arnoldi , being always much over body length; the wings are narrower (in a female of S. gracile they measure 14 by 4 mm.) and much less infuscated; the third segment of the antennae is decidedly shorter (in S. gracile it is about twice as long as wide at base and amply as long as the two basal segments together).

Stenobasipteron difficile sp. nov.

Type female from Cloudlands, 6,000 ft., Vumbu Mts., Southern Rhodesia, 6 to 17 April, 1923 (without collector). In the collection of the Rhodesia Museum.

A medium-sized, rather thickset, black species, covered with cinereous tomentum on the under side, with a black bloom on the upper side, the dorsal surface of the abdomen somewhat shiny. Antennae and legs reddish brown; the last antennal segment and the hind tarsi darker. Pilosity moderately long and dense on head, thorax and base of abdomen, yellowish white

12

Psyche

[February

ventrally, pale russet dorsally. Proboscis about as long as the body. Wings moderately long, uniformly smoky all over.

Female. Integument black even on the face; scutellum clove-brown. Two basal segments of antennae reddish clove- brown; the last segment much darker; the style black. Palpi clove-brown. Proboscis brownish black. Legs reddish clove- brown; the tip of the tibiae and the tarsi more infuscated; the hind tarsi almost black; claws black.

Pilosity rather long and abundant on head, thorax, and base of abdomen. Vertex, front, and face with numerous, erect, black hairs; cheeks and posterior orbits with a long and dense beard of greyish white pile. Dorsum of thorax and scutellum uniformly covered with loose, erect, moderately long, reddish yellow pile, more russet on the scutellum; sides and ventral face with long and dense, greyish white hairs, more yellowish below the wings. Dorsally the abdomen bears on the first and basal half of second tergites long, erect, reddish yellow pile, similar to that of the thoracic dorsum; the remainder of the dorsal side bears many scattered, short and more or less appressed, black hairs; and in addition a very sparse, long, erect, black pilosity; ventrally there is on the sides a dense, yellowish white, matted pile, but the largest part of the sternites has but a very few, short, appressed, white hairs. Coxae and femora with long, erect, greyish white pile; the hairs on the tibiae and tarsi very short and black. The ground color of the integument on the ventral side of thorax and abdomen is completely hidden by a dull, cinereous white bloom. On head and dorsum of thorax the pruinescence is dull and very dark brown, somewhat more cinereous on the sides of the dorsum and on certain areas of front and face. The sides of the face rather shiny. On the dorsal face of the abdomen the pruinescence is velvety black, with a somewhat oily sheen, and there are two transverse, ill- defined spots of a yellowish grey, dull bloom near the anterior margin of the second, third and fourth tergites, so that the abdomen appears quite distinctly spotted.

Head moderately flattened, much broader than the thorax; semi-elliptical seen from above; in profile the face ts moderately projecting, much less so than in S. arnoldi] when seen in front

1925] New N emeslrinidce from Rhodesia and New Guinea

13

the head is kidney-shaped and nearly twice as wide as high in the middle. Front rather wide, broadest at the insertion of the antennae where it measures a little over half the width of the eye; inner orbits moderately converging toward the anterior ocellus, where the front is slightly over half as wide as at the antennae. Sides of the vertex parallel. Ocellar tubercle short and flat, with a slight transverse depression below its middle, hardly separated by a notch from the inner orbits; ocelli placed in a short isos- celes triangle, the posterior ocelli distinctly, but slightly, closer to each other than to the anterior ocellqs; anterior ocellus larger than, though not quite twice the size of, a posterior ocellus, short elliptical, occupying a little less than one-third of the width of the front. Eyes bare. Antennae (Fig. 16) short, small, placed on the sides of the face, close to the inner orbits; basal segment cylindrical, nearly one and a half times as long as wide, squarely truncate at apex; second segment but little shorter than the first, slightly longer than wide, broadly truncate at apex; third segment much flattened, very elongate pear-shaped, over twice as long as wide, much longer than the two basal segments together, widest in its basal half and thence rather rapidly tapering to the straightly truncate, narrow apex. Style about the length of the whole antenna, only two-jointed; the basal division a little shorter than the second antennal segment. Front very feebly convex between anterior ocellus and antennae. Face moderately swollen, forming a low, blunt cone, without grooves. The lower portion of the head is but slightly and very broadly excavated in the middle between the cheeks. Proboscis very long and slender, reaching about the tip of the abdomen when folded beneath the body; in the specimen in hand it is directed vertically downward, with the apical third curved forward; labella very thin and elongate. Palpi short and slender, three-jointed; the two apical segments much longer and of about equal length; the third obtuse at apex. Body rather thickset. Thorax about as wide as thick; dorsum nearly square; trans- verse suture deep on the sides over about one-third of the width of dorsum, continued backward to near the scutellum. Scu- tellum large, semi-elliptical, its posterior margin distinctly separated from the disk by an impressed line. Abdomen flat-

14

Psyche

[February

tened dorsal ly, much wider than the thorax; the four basal segments together much shorter than wide; the succeeding, apical segments much narrower, gradually decreasing in width, partly retractile as a telescope-shaped ovipositor. The last segment ends in two comparatively wide, long, straight, bluntly pointed lamellae. Legs moderately heavy; femora slightly thickened, more distinctly swollen on the forelegs.

Wings moderately long and narrow, not quite four times as long as wide. Costal margin nearly straight; the posterior margin gradually narrowed in its basal quarter, where a narrow, but distinct alula extends from the axillary excision to the base of the wing. Wings distinctly and uniformly smoky; veins dark clove-brown; the epaulets and basicosta black; the epaulet with a few black hairs. Venation of the usual type of the genus; fourth posterior cell with a short petiole at the base; sixth longi- tudinal vein very slightly undulate (nearly as in arnoldi); no “bulla” at base of second vein; axilla^ vein not developed beyond axillary incision.

Length not including ovipositor (to apex of tergite 4), 11.5 mm.; greatest width of abdomen, 6.5 mm.; length of proboscis, 11.5 mm.; length of wing, 15 mm.; width of wing, 4 mm.

This species is exceedingly close to Stenobasipteron gracile Lichtwardt, much more so than S. arnoldi. There are, however, a number of differences: the body is more thickset; the legs are stouter; the anterior ocellus is smaller, not quite twice the size of a posterior ocellus; the third antennal segment has a different shape; the style is only two-jointed; the sixth longitudinal vein is quite straight; the lamellae of the ovipositor are broader, etc. It is difficult to believe that these discrepancies are all due to individual variation. Moreover, the unknown male may show further characters.

Stenobasipteron gracile Lichtwardt

The original description of this species (Deutsch. Ent. Zeitschr., 1910, p. 615) is extremely brief and is reproduced here for the benefit of Rhodesian entomologists. It is said to be so

1925] New N emestrinidce from Rhodesia and New Guinea

15

similar to S. wiedemanni Lichtwardt, that only the differences are noted: .“Smaller and more elegant in the whole build of the body; the color is markedly paler than in S. wiedemanni and has a more greyish tinge; while the uniformly colored surface of the wing is also more smoky grey, showing but a narrow, yel- lowish-brown stripe along the anterior margin. Sharp dif- ferences are the absence of the “bulla’’ in the wing of both sexes; the inequality in size of the ocelli, of which the anterior one is twice as large as one of those placed at the occiput; and the bud-like, rounded shape of the male hypopygium, which is larger in proportion to the size of the animal. Length of the body, 13 mm.; of the proboscis, 10 mm.; of the wing, 15 mm.” This description was drawn on a male and female from Mazoe, Mash- onaland (Southern Rhodesia), in the British Museum. Later,, Lichtwardt recorded as S. gracile two females from Barberton,, Transvaal, in the South African Museum (Entom. Mitteil. Berlin, IX, 1920, p. 97). Bezzi (Ann. South African Mus., XIX, 1924, p. 171) does not appear to have seen it.

'I have referred to S. gracile one female and two males of Cloudlands, 6,000 ft., Vumbu Mts., Southern Rhodesia, 6 to 17 April, 1923 (without collector). It must be stated, however, that the description quoted above does not allow a positive identification, so that I feel justified in giving some additional data that might help in separating S. difficile from what I take to be S. gracile.

The measurements of the three specimens before me are as follows:

Female. Length not including ovipositor (to apex of tergite 4), 11.5 mm.; greatest width of abdomen, 5.7 mm.; length of proboscis, 10.5 mm.; length of wing, 14 mm.; width of wing, 4 mm.

Males. Total length, 11 and 11 mm.; greatest width of abdomen, 5.5 and 5 mm.; length of proboscis, 11 and 10 mm. length of wing, 13 and 12.5 mm.; width of wing, 4.2 and 4 mm.

In these examples the proboscis may therefore be said to be about as long as the body. Quite apart from individual variation, a certain latitude should be allowed in judging these relative lengths, as the body length certainly changes after death, while

16

Psyche

[February

the proboscis is to some extent retractile. Lichtwardt’s meas- urements seem to indicate that in his specimens the proboscis was considerably shorter than the body, but this may be de- ceptive. It is not stated whether the measurements referred to the male or to the female, and, if the latter was measured, whether the body length includes the ovipositor.

In the Cloudlands female the integument is generally black; face, antennae, palpi, and legs rather bright reddish clove-brown; the femora more yellowish brown. The long pilosity is greyish white ventrally; black dorsally, even on the dorsum of thorax and scutellum; on the dorsum of the abdomen there is a mixture of a few, shorter, white hairs. The pruinescence is dull all over, cinereous white ventrally, very dark greyish brown dorsally. The abdomen is not spotted. Face more prominent than in S. difficile, more as in S. arnoldi. Ocelli in a short isosceles triangle, as in S. difficile, but the anterior ocellus is larger, being dis- tinctly twice the size of a posterior ocellus and occupying a little more than one third of the width of the front. The antennae (Fig. lc) are shaped much as in S. difficile, the third segment being slender, pear-shaped and over twice as long as wide at base; but the arista is considerably longer than the whole an- tenna and three-jointed. The legs are relatively thinner and the lamellae which terminate the ovipositor narrower than in S. difficile. The sixth longitudinal vein is perfectly straight before the apical curve.

The two males are structurally alike and differ mainly from the female in the usual sexual peculiarities (wing much widened at anterior margin, beyond the middle; front much narrowed above so that the anterior ocellus occupies nearly the whole width; abdomen ending in a bluntly swollen hypopygium). The integument of scutellum and dorsal side of abdomen is to a large extent clove-brown. The long, black pilosity of dorsum of head and thorax shows a tendency to be russet brown, especially on the front and the scutellum. The third antennal segment is a little shorter than in the female, but still at least twice as long as wide at base.

1925] New N emestrinidce from Rhodesia and New Guinea

17

Key to the Known Species of Stenobasipteron.

1. Wings relatively short (7.5 mm.); with a short, but chitin-

ized axillary vein, bent at an angle in the middle. Proboscis shorter than body (4.5 mm.). Ocelli of same size, in an equilateral triangle. Front (?) a little narrower than

one eye. Small species (7 mm.) ( c? unknown)

S. minimum Bezzi.

Wings much longer than body, with the axillary vein hardly distinguishable, not chitinized, straight. Front of female much narrower than one eye. Larger species 2 .

2. First basal cell with a “bulla” near the base of second longi-

tudinal vein.] Ocelli of nearly same size, in an isosceles triangle. Style of antenna three-jointed. Proboscis much longer than the body (24 mm.). Large species (17 mm.)

S. wiedemanni Lichtwardt. First basal cell without “bulla.” Medium-sized species (10.5 to 13 mm.) 3 .

3. Proboscis much longer than the body (14.5 to 16 mm.).

Third antennal segment short, slightly over one and one- half times as long as wide; style three-jointed, about twice the length of the antenna. Length (9): 10.5 to 11 mm.

(cP unknown). S. arnoldi, sp. nov.

Proboscis about as long as the body or a little shorter. Third antennal segment over twice as long as wide at base 4 .

4. Style two-jointed, about as long as the whole antenna.

Anterior ocellus (?) occupying a little less than one-third of the width of the front. Length ($): 11.5 mm. ( c P

unknown) 'S. difficile, sp. nov.

Style three-jointed, much longer than the whole antenna. Anterior ocellus ( $ ) occupying a little over one-third of the width of the front. Length ( $ , without ovipositor) : 11.5 mm.; (cT): 11 mm S. gracile Lichtwardt.

1This bulla probably corresponds to the minute swellings of the wing membrane known as “nygmata” in certain Neuroptera, Trichoptera, Panor- pata, and Hymenoptera. They are apparently not known in other Diptera. See W. T. M. Forbes, Ent. News, XXXV, 1924, pp. 230-232, PI. V.

18

Psyche

[February

Nycterimyia Lichtwardt

This extraordinary genus of flies is at present known in seven species: N. dohrni (Wandolleck) of Sumatra, Mafor,1 and the Andaman Islands; N. horni Lichtwardt, of Northern Queensland; N. kerteszi Lichtwardt, N. fenestro-clatrata, and N. fenestro-inornata Lichtwardt, of Formosa; N. capensis Bezzi, of Natal; and the New Guinean species described below. Although all species are closely allied, the distribution of the genus is extremely discontinuous.

The structure of the antennae appears to be quite different in N. papuana from what has been described in other species. Of A. dohrni , Wandolleck (Entom. Nachricht., XXIII, 1897, p. 251) wrote originally: “Fuhler 3-gliedrig, gelb, drittes Glied stabformig mit welligen Conturen; and der Spitze tragt es ein ganz kurzes, feines, dqrchsichtiges Tastharchen.” Lichtwardt (Deutsch. Entom. Zeitschr., 1909, p. 647) says of the same species: “An den Fiihlern ist das dritte Glied nach vorn ver- breitert und mit einer stiftartigen, starken, apikalen Borste versehen.” Bezzi (Ann. South African Mus., XIX, 1924, p. 169) describes the antennae of N. capensis as “very short, with the third joint rounded and smaller than the preceding one; they are pale yellowish like the rather thick style, which is twice as long as the antenna/’ In my example of N. papuana (Fig. 2a), the third joint is extremely slender and ends in a narrower, seta-like portion, although no trace of suture could be discovered between the basal and apical sections. It agrees therefore best with Wandolleck’s account, but I can not find a differentiated tactile hair at the tip. I am inclined to believe that the “thick style” in Bezzi’s description of A. capensis represents the true third antennal segment, while his “third joint” is what I describe as the second segment.

Nycterimyia papuana sp. nov.

Type male from “Baie du Geelvink, New Guinea,” (Raffray and Maindron Coll., 1878). In the collection of the Paris Mu- seum.

xThis appears to be a misspelling for Mapor, one of the Rhio Islands, between Singapore and Sumatra.

1925] New N emestrinidce from Rhodesia and New Guinea 19

A medium-sized, robust, brown black species, covered with a dull, reddish brown tomentum; legs and antennae testaceous. Pilosity brownish grey on head and thorax; abdomen almost destitute of hairs. Wings long, deeply bisinuate along the posterior margin, deep reddish brown; an elongate and narrow hyaline streak in the fourth posterior cell and faint indications of hyaline in the center of the combined first and second posterior and of the second basal cells.

Male. Integument apparently black, though the body is so uniformly covered with tomentum that it is difficult to see the proper color. Antennae and legs pale testaceous; coxae more brownish; apical half of claws brownish black.

Head and thorax with abundant, long, erect, brownish grey pilosity, which is denser on the ventral side. Hairs of the abdomen very short and sparse, dark grey; somewhat more abundant and longer ventrally and on the sides of the second tergite. Coxae and femora with moderately long, reddish grey hairs; the pilosity of tibiae and tarsi much shorter, but of the same color. Head, thorax, and abdomen are covered with a dull, cinnamon red bloom. There are no traces of dull stripes on the thorax nor of spots on the abdomen; but the second tergite bears close to its base a deep, transverse groove, which is shiny except on the middle; in addition there is a short, transverse, shiny depression on the side of each of the tergites 2, 3, 4, and 5.

Head large, much flattened, a little broader than the thorax; semi-elliptical in profile and from above; kidney-shaped and nearly twice as wide as high when seen in profile. Front nar- rowly triangular, widest at the antennae where it measures about one-half the width of the eye; the inner orbits strongly converging above, where they come extremely close together for a short distance below the anterior ocellus, though not actually touching. Vertex triangular. Ocellar protuberance quite prominent, short, deeply divided behind from the inner orbits which project a considerable distance beyond the occipital margin of the vertex. Ocelli large, of about the same size, placed in an equilateral triangle. Eyes bare, composed in their upper half of large facets which gradually merge into the much smaller ommatidia of the lower half. Antennse (Fig. 2a) very small, placed a short dis-

20

Psyche

[February

tance from the inner orbits on the upper portion of the curved slope which leads into the deep transverse depression that separates the front from the face; basal segment short, much thicker than long, widened and crescent-shaped at the apex; second segment disk-shaped, almost circular from the side, as long as the first; third segment apparently fused with the style, the whole being over twice the length of the two basal segments together, extremely slender and narrow, strongly tapering from the basal third to the apex which is very' sharply pointed Front regularly curved from vertex to antennae, below which it droops deeply into a very pronounced transverse groove sep-

Fig. 2. Nyclerimyia papuana. a, right antenna drawn from the inner side; b, wing.

arating the face. Face sunken between the eyes, the median, shorter portion sharply divided from the lateral areas by deep, vertical grooves. Palpi and proboscis if present, extremely reduced and not to be distinguished among the long pilosity. Body broad and thickset. Thorax about as thick as, but shorter than, wide; its dorsum distinctly convex. Transverse suture well-marked on the sides over less than one-third of the width of the dorsum, obliquely continued behind to a short distance from the scutellum. Scutellum large, nearly elliptical, its posterior margin not separated from the disk. Abdomen broad and short, but little longer than wide, quite convex dorsally and somewhat curved down at the apex. The several segments are distinctly constricted, their apical portion being slightly swollen. The first tergite is very short and mostly covered by the scutellum. Second tergite much the longest, about as long as the two following tergites together; in its basal half it is broadly grooved transversally, the bottom of the channel being

1925] New Nemestrinidce from Rhodesia and New Guinea 21

shiny (except medially) and finely alutaceous; in addition there is on the posterior third of the tergite on each side a short, narrow and rather shallow, transverse groove, where the integument is also shiny and alutaceous. Third, fourth, and fifth tergites of about equal length, each on the sides, shortly behind the an- terior margin, with a short, transverse shiny groove similar to that found on the hind third of the second tergite. The apical tergites are much shortened and somewhat retracted ventrally; they end in a prominent knob containing the large genitalia. Legs long and stout; the hind legs considerably longer than the anterior and middle pair. Fore and mid femora moderately and rather uniformly swollen, much thicker than the tibiae; hind femora elongate club-shaped, distinctly swollen toward the apex. Tibiae slender, not appreciably thickened at the apex. Tarsi short, narrower than the tibiae.

Wings (Fig. 26) long and moderately wide, over three times as long as the greatest width, which lies at the apex of the anal cell. The fore margin quite straight; the hind margin wavy between the tip of the fifth longitudinal vein and the apex of the wing: of the two, deep sinuations the proximal one, between the tips of the diagonal and fifth veins, is much the longest. Alula small, but quite well developed (as figured by Lichtwardt for N. horni and allies). Epaulet and basicosta clove-brown. Wings of a deep brown, opaque color, with a distinct cinnamon red tinge. A whitish hyaline, narrow, somewhat curved, longi- tudinal streak, with a pearly sheen, occupies the center of the fourth posterior cell (the cell immediately below the discal1); it begins quite a distance from the base of the cell, where it is widest, and gradually tapers to a short distance from the diagonal vein. There are no other well-marked hyaline spots; but the center of the combined first and second posterior and of the second basal cells is distinctly subhyaline and there is even a faint indication of a hyaline area in the second basal cell. The two wings are exactly alike in this respect. Veins bright reddish

xIn Nycterimyia there are only four posterior cells differentiated. To make the nomenclature of the wing homologous with that of the majority of Nemestrinidae which have five posterior cells, it is necessary to assume that the first and second are fused; the cell here called the fourth then corresponds to the cell of the same name in Prosoeca, for example.

22

Psyche

[February

brown, darker basally. Venation as in the other species of the genus: in details it agrees best with Lichtwardt’s figure of V. kerteszi (Entom. Mitteil., I, 1912, Pl. II, fig. 2), but the short cross-vein which unites the first and second' longitudinals is much farther removed from the long cross-vein connecting the second and third longitudinals. It should also be noted that the auxiliary vein (or subcosta), which both Wandolleck and Lichtwardt figure as uniting with the first longitudinal about the middle of the wing, really continues its course independently to near the base, as in other Nemestrinidae'; furthermore it is connected, a short distance from the base, with the costa by a humeral cross-vein, apparently overlooked by these authors. The apex of the discal cell is far removed from the base of the combined first and second posterior cells. The costa extends to beyond the tip of the fourth longitudinal vein, whence it gradual- ly fades away to the apex of the wing.

Total length, 11 mm.; greatest width of abdomen, 5 mm.; length of wing, 13.5 mm.; greatest width of wing, 4 mm.

The species is closely allied to N. dohrni and N. horni , but differs conspicuously in the markings of the wing.

1925]

Some Species of the Genus Leucospis

23

SOME SPECIES OF THE GENUS LEUCOSPIS.1

By Charles T. Brues.

The recent paper by Mrs. Weld2 has encouraged me to undertake the identification of the species of Leucospis which have accumulated in my collection. Among these are several undescribed forms of which descriptions are given below, to- gether with a few notes on known species, mainly in regard to their geographical distribution.

Leucospis birkmani sp. nov.

$ . Length 12.5 mm. Black, with yellow and ferruginous markings and strong metallic reflections. The yellow is dis- tributed as follows: antennal scape; median round spot near anterior margin of pronotum; entire posterior and lateral mar- gins of pronotum; transverse band on mesonotum just before base of scutellum; four anterior knees; triangular spot at base of hind femur below and band along apical two thirds of upper edge; outer two thirds of hind tibia and extreme outer tip of hind coxa. The ferruginous markings include the base of an- tennal scape; tegulse; inner side of front femora and their tibiae and tarsi; middle legs; hind coxae, except lower surface; inner margin of hind tibiae and their tarsi entirely; sides of propodeum; middle of first abdominal segment above and the elongate ventral plates of the abdomen. Face finely punctate reticulate; inter-antennal projection with a strong median carina; nar- rowest width of face clearly less than its height; malar space .one-third longer than the second flagellear joint; ocelli on a distinctly elevated tubercle, the paired ones as close to one another as to the eye margin; occipital margin prominent medially, as high as thd ocellar tubercle; vertex rather finely punctate, confluently so in front; occiput irregularly reticulate, aciculate only near the middle. Antennae with the second and

Contribution from the Entomological Laboratory, Bussey Institution, Harvard University, No. 246.

Clara Jamieson Weld, Studies on Chalcid-flies of the Subfamily Leucos- pidinae, with Descriptions of New Species. Proc. U. S. Nat. Mus. Washington, vol. 61, art. 6, 43 pp., 1922.

24

Psyche

[February

third flagellar joints clearly longer than wide, those beyond becoming quadrate and then distinctly transverse; inner eye- margin not emarginate. Pronotum without trace of a transverse carina, its surface densely, coarsely, separately punctate, the posterior yellow band smooth, except at the sides; mesonotum more irregularly so, with some of the punctures confluent, especially at the sides ; scutellum strongly convex, more shining, with well separated but shallower punctures; post- scutellum broadly crescentic, coarsely reticulate with a smooth raised margin. Propodeum medially as long as the postscutellum, with a median carina that is strongly raised behind; lateral and apical carina distinct, the surface between reticulate coarsely at the sides, very finely near the middle. First abdominal segment scarcely one-half longer than wide, narrower toward base, its dorsal grooves broad and shallow, the sides coarsely sparsely punctate at base and very densely and finely so at apex; abdomen considerably widened beyond the first segment, the fifth (the one preceding the base of the ovipositor) fully one half wider than the first; apex narrowly rounded; fourth segment with mod- erately coarse, well separated punctures at the base, becoming shagreened on the apical half; fifth coarsely punctured on its basal three fourths, with the apex very finely punctate or sha- greened; following segments coarsely punctate, much more closely so above. Ovipositor reaching the tip of the scutellum. Propleurse finely closely punctate below, shining and obsoletely punctate reticulate above; mesopleura, metapleura and sides of propodeum increasingly more coarsely and sparsely punctate. Hind coxae entirely shining, finely and densely punctate on the underside; sides below confluently punctate, above with separate, punctures and with a small smooth space near the upper angle; upper edge thin and sharp behind, without tooth. Hind femur shining, very minutely and closely punctate above, more coarsely and sparsely below, the punctures widely separated on the lower edge; length distinctly more than twice the width; basal tooth the largest, but not conspicuously enlarged, followed by nine much smaller ones of which those near the basal tooth and at apex are more minute than the intermediate ones. Anterior wings distinctly infuscated on their anterior half.

1925]

Some Species of the Genus Leucospis

25

The metallic reflections on the body are mainly green, noticeable on the head in the antennal grooves, vertex and occiput, on the under surface of middle and hind coxae, the apices of first, third and fifth abdominal segments and upper side of propodeum. On the face and front of vertex there are some purplish reflections.

Type from Fedor, Lee County, Texas (Birkman). The species is named for its discoverer, the Rev. G. Birkman whose collections have added greatly to the knowledge of the insect fauna of this portion of Texas.

This species will run to L. cayennensis Westw. in Schletterer’s key (foe. cit.) and also in the one given by Weld (foe . cit., p. 8) but differs by the punctate hind coxae, distinctly clavate ab- domen and maculate thorax. From L. distinguenda Schletterer, the much longer ovipositor will serve readily to distinguish it.

It is a surprise to find this large fine species undescribed and I had thought that it might be regarded as a northern subspecies of L. cayennensis, but there are so many striking differences that it can hardly be regarded as such.

Leucospis muiri sp. nov.

dL Length 8 mm. Black, with yellow markings and some fulvous or ferruginous ornamentation. The lemon-yellow is as follows; scape of antenna; a broad transverse band on prono- tum, curving forward and attaining the anterior margin at the sides; a streak above each tegula and a pair of discal spots on mesonotum; scutellum, except large triangular spot medially in front; transverse streak on postscutellum ; large spot above on mesopleura; large triangular one above hind coxa; first ab- dominal segment above, except basally and at sides before apex; two transverse bands on gaster, the first quite narrow. The tegulae and second and third joints of antennae are ferru- ginous and the body spots are more or less margined with fer- ruginous. Legs blackish basally but yellow and ferruginous beyond; the outer tips of four anterior femora and lower edge of hind femur yellow; remainder of femora and tarsi entirely ferru- ginous, except that the hind femur is black along the middle and

26

Psyche

[February

its tibia is streaked with black internally. Wings rather deeply infuscated, except at base. Face finely vertically shagreened or punctulate with a short median carina below the antennae; malar space as long as the second flagellar joint; antennal cavities strongly transversely striate; vertex densely and rather finely punctate, the posterior ocelli but little further from one another than from the eye : occipital carina sharp medially, coming very close to the ocelli; occiput finely circularly striate-punctate and shining. Antennae with all the flagellar joints decidedly longer than thick. Frothorax, mesonotum and scutellum rather finely and very densely but not confluently punctate, less distinctly shining than usual; pronotum with a single transverse carina medially close to the posterior margin, unusually long, three- fourths the length of the mesonotum. Scutellum oval, with the hind edge distinctly margined. Postscutellum somewhat promi- nent, semicircular, the crenate margin with a median emar- gination. Propodeum very coarsely rugose with a strong,, almost dentiform median carina at each side of which lies a less prominent carinate line; sides also distinctly carinate. Pleurae punctate, much more sparsely and coarsely so behind. Abdomen short and very strongly clavate, the gaster fully thrice as wide as the petiole. Petiole slightly wider than long, its sides parallel; highly convex above and armed below near apex with a cons- picuous long slender erect tooth; abdomen shining at base, more opaque apically, closely punctate, the punctures elongated and giving the appearance of a longitudinal trend to the sculp- ture. Hind coxae very finely and densely punctate below and inwardly above, the punctures becoming very sparse and the surface shining outwardly above; upper edge without tooth and broadly rounded, not sharply ridged as in most species. Hind femur very minutely punctate; with a large triangular tooth at the middle, followed by three widely spaced small teeth, followed by five or six still smaller ones becoming minute at the apex of the femur.

Type from Laloki, Papua, 1910 (F. Muir). The body is not conspicuously pubescent, although the face and abdomen are clothed with short, pale glistening hairs. There is no trace of metallic color on the body.

1925]

Some Species of the Genus Leucopsis

27

This species is similar to L. mysolica Kirby1 in color and in the dentition of the hind femora. It may be separated readily by the sculpture of the vertex, position of ocelli, and form of the antennae, as well as by the single carina on the pronotum.

Leucospis malabarensis sp. nov.

$ . Length 9 mm. Black, with yellow ornamentation and some ferruginous markings, without metallic color. The yellow markings are as follows; antennal scape below, oval spot on each frontal prominence, as long as the scape; two narrow trans- verse bands on pronotum, the anterior one curved forwards laterally and the posterior one not reaching the sides; a thin streak above the tegulse and a pair of small spots on middle of mesonotum; narrow arcuate band on posterior margin of scu- tellum; large triangular mark below tegula; short streak on metapleura above; spot at upper angle of hind coxa; pair of broad lateral stripes on basal half of first abdominal segment, their bases nearer to the median line; narrow band at base of fourth segment, extending halfway down the side; broader, complete apical band on fifth and a pair of short vertical lines just before tip of abdomen; margin of hind femur, except the toothed portion; small elongate spot just above apical teeth; anterior knees, external streak on all tibiae, not attaining the base on the hind pair. Tegulae, apices of all coxae and more or less of fore and middle femora and of all tibiae rufopiceous; tarsi ferruginous. Antennae more or less rufous; wings moderately infuscated, except at base. Face microscopically reticulate punctate, external margin of antennal cavity distinctly carinate; punctures of vertex moderately large and well separated; post- erior ocelli almost twice as far from each other as from the eye margin; occipital carina rather weak; occiput shining, dis- tinctly striate only at the middle; malar space slightly longer than the second flagellar joint. First three joints of flagellum longer than wide, those beyond quadrate. Pronotum three-

tJourn. Linn. Soc. vol. 17, p. 69 (1882); cf. also Schletterer, Berliner Entom. Zeits., vol. 35, p. 236 (1890) and Enderlein Arch. f. Naturg., Jahrg. 67, vol. 1, p. 216 (1901).

28

Psyche

[February

fourths as long as the mesonotum, with a single carina near the posterior margin which is also quite distinctly carinate; surface shining, with the punctures rather small and closely placed; mesonotum behind and the scutellum more coarsely punctured; postscutellum short, transverse, simple. Propodeum short, finely reticulated, not carinate except for the lateral carinse which are very distinct although not prominent. Abdomen shining, its punctures not densely placed except at apex; first segment two-fifths the length of the abdomen, twice as long as wide and distinctly broadened at the middle. Ovipositor very long, reaching well beyond the apex of the scutellum. Propleura rather weakly confluently punctate; mesopleura and metapleura more coarsely so, especially the metapleura where the punctures become confluent above. Hind coxa finely densely punctate below, very sparsely above where the surface is highly polished; upper edge very sharp behind, but without tooth. Hind femur broad, including the teeth scarcely twice as long as wide; the surface shining and finely, evenly punctate; basal tooth very small; three succeeding ones long and widely spaced, the third broadest and blunt at apex; following tooth close and somewhat shorter followed by several closely crowded ones that become rapidly shorter. Body with a moderately dense coat of short white pubescence longer on the pleurae and especially on the sides of the propodeum.

Type from North Malabar, Southern India (A. P. Nathan).

In Schletterer’s key ( loc . cit ., p. 167) this species will run to L. japonica Walker from which it differs in the type of den- tition of the hind femora. In general appearance and in denti- tion of the hind femora it is similar to L. macrodon Schletterer, but the first abdominal segment is much narrower and longer, the ovipositor is longer and the ocelli much more widely sep- arated. L. macrodon is very variable in color, but the spot is reduced only in very dark specimens. From L. quettaensis Cam. and L. nursei Cam. both from Baluchistan, this species differs conspicuously in color and in the simple postscutellum. L. viridissima Enderl. from Ceylon is entirely different from the present form.

1925] Some Species of the Genus Leucospis 29

Leucospis japonica Walker.

Notes on Chalcidise, pt. IV, p. 56 (1871).

This species occurs also in China from whence I have a specimen collected by N. Gist Gee at Soochow.

Leucospis affinis Say.

A female from Jacumba, California (W. M. Wheeler) has the pale whitish ornamentation of L. hicincta Viereck. The hind femora have the basal and apical pale areas connected by a pale band below, a condition which seems never to occur in individuals from the eastern states.

30

Psyche

[February

THE EFFICIENCY OF BIRDS IN DESTROYING OVER- WINTERING LARVAE OF THE EUROPEAN CORN BORER IN NEW ENGLAND.1

By Geo. W. Barber.

Cereal and Forage Crop Insect Investigations, Bureau of En- tomology, U. S. Dept, of Agriculture.

Several years ago, not long after investigations of the European corn borer ( Pyrausta nubilalis Hiibn.) were begun by the Bureau of Entomology of the United States Department of Agriculture, it was frequently observed that cornstalks infested by the larvae of this insect showed in the spring of the year numerous holes along the stalks, the burrows of the insect beneath these holes being empty. This was the first evidence of any appreciable feeding by birds on this insect. Such evidence of bird feeding has been found each spring and it is now possible to associate this work with the downy woodpecker ( Dryobates pubescens medianus Swainson) a winter resident in this region. In numerous instances, this bird has been observed at close range at work on the infested standing cornstalks. Plate 1a, shows sections of cornstalks from which the larvae of the corn borer have been removed by this bird. This type of feeding by chickadees ( Penthestes atricapillus atricapillus Linn.) has also been observed by Mr. F. H. Mosher.

Within the last few years observations have shown another type of feeding by birds on the overwintering larvae of this insect. This is the shredding of cornstalks illustrated in Figure IB, and is the result of feeding by grackles, blackbirds, starlings and probably several other species of migrating birds. These birds arrive in the latitude of Boston, Mass., from the middle to the last of April. Such work was especially noticeable in corn- stalks that had been piled in the fall or in stalks that had fallen over for one reason ojr another and lay on the .surface of the soil. These birds have frequently been observed feeding in flocks in the spring, and in a short time they are able to gather the larvae

Contribution from the Bureau of Entomology, U. S. Department of Agriculture in cooperation with the Entomological Laboratory of the Bussey Institution, Harvard University, Bussey Institution No. 248.

1925] Efficiency of Birds in Destroying Larvce of Corn Borer 31

from quite a number of cornstalks. They are able also to shred infested corn stubble and take the larvae in the more exposed positions, but apparently are not able to reach the larvae con- tained in standing stalks.

In the fall of 1922, when it was apparent that the birds were becoming a really important factor in the reduction of the num- bers of the corn borer, experiments were undertaken to determine how extensive such feeding was. These experiments were also carried on during the winter of 1923-1924. The object of this work was to obtain information on the extent of the combined feeding by all species of birds concerned rather than the extent of feeding of any particular species, the intention being to obtain as far as possible a picture of the present importance of birds as a group in relation to this insect rather than a study of the value of any one particular species.

The studies pursued during the fall of 1922 and the spring of 1923 may be treated under two heads; first, the extent of bird feeding on the larvae in infested cornstalks placed in the field for this purpose; second, the extent of feeding by birds on larvae in host plants that remained undisturbed in natural positions during the winter.

In the first part of this work twenty representative locations were selected throughout the infested area of eastern New Eng- land. In each of these locations ten stakes were set upright in the soil, there being six infested stalks fastened to each of these stakes. The three following types of corn were represented at each location: pop corn, sweet corn (Golden Bantam) and field corn (Longfellow Flint). Counts of the larval population of representative stalks during the fall gave an average figure as to the number of larvae expected from each stalk, and from this average the number of larvae expected from each station was computed. This series of experiments was placed in the field in November, 1922, after all larval activity had ceased, and the stalks were collected in April, 1923, before larval activity had commenced in the spring. The chance of losing an appreciable number of larvae by migration was small, since during this period the larvae were entirely dormant and inactive. Of these twenty experiments five showed extensive feeding by birds

32

Psyche

[February

when examined in the spring, mostly the work of woodpeckers. Of the remaining fifteen experiments one was destroyed by an over-anxious farmer, one was partly destroyed by a tractor and the others showed only very slight traces of feeding by birds or no evidence of bird feeding whatever. Table Number 1 shows the extent of bird feeding on the five experiments attacked, and the recovery of larvae from stations that escaped noticeable bird feeding. The average recovery of larvae from experiments not attacked by birds was 1,090 larvae per station as compared with the average expectancy of 1,223.2. This apparent loss of 10.8 per cent of expected larvae per station was undoubtedly

TABLE I

Extent of Feeding by Birds on Experimental Material in the Spring of 1923.

Experiments on which birds fed.

		<v		1		>.1 §11	- .b 0 a <v
Location	e placet 1922	recover 1923	. larvse pected	> £ U <D 43 > 0	3parent loss	0^ <0 C Id c	Jv CU S- O U 4, O » a- I’S
	d Q	<D CTj Q	0 x £	6 cj £	<	b -b $ 1 1 > § Oh <j ctJ 0	Cb rt.S m 0
Scituate, Mass.	XI-3	IV- 1 1	906	123	783	86%	84%
Newbury, Mass.	XI-10	IV- 1 4	1289	214	1075	83%	81%
Rockport, Mass.	XI-10	1V-I2	1289	209	1080	83%	8l%
Medford, Mass.	X-25	IV-9	1289	390	899	69%	66%
Arlington, Mass.	XI-2	IV-9	809	576	233	27%	12%

Experiments on which birds did not feed.

I

Bristol, N. H.	XI-8	1V-29	1289	1397
Framington,N.H.	XI-7	IV-29	1289	1341
Wells, Me.	XI-7	IV-28	1289	1132
Concord, Mass.	XI-4	IV-io	809	854
Falmouth, Mass.	X-31	IV- 7	1289	1135
Quincy, Mass.	XI-4	IV- 1 1	906	944
Harwich, Mass.	XI-i	IV- 6	1289	„ 909
Methuen, Mass.	XI-2	1V-14	1289	T014
Manomet, Mass.	X-30	iV-6	1289	1081
Tyngsboro, Mass.	XI 8	[V-14	1289	990
Wareham, Mass.	Xf-I	IV-8	1289	io56
Wellfleet, Mass.	X-31	IV- 7	1289	1106
Worcester, Mass.	X-26	IV-II	1289	1222

1925] Efficiency of Birds in Destroying Larvce of Corn Borer 33

caused by migration of a few larvae and the loss of small pieces of stalks containing larvae during transportation of the corn- stalks used in the experiment.

The average winter mortality in the 18 experiments listed in table No. 1 was 10.5% percent. The average per cent of larvae credited to bird feeding in the five stations where stalks were attacked, was 61. The average per cent of larvae credited to bird feeding in the 18 stations recovered was 17.

The stations that showed extensive feeding by birds (Fig. 1) are all within the area most heavily infested by the European corn borer and localities where infestation has been severe for several years. Because of this fact and because no marked evidence of bird feeding was found in areas slightly infested or areas that had become heavily infested by the insect within the last year or two, it would appear that woodpeckers are aware of the fact that infested cornstalks contain desirable food only in this heavily infested area, and that in more sparsely infested regions or in areas where infestation had but recently become severe they are for the most part still unfamiliar with the exis- tence of this source of food.

Several of the more commonly infested weeds and cul- tivated plants were also tied to stakes to observe possible feeding by birds on larvse contained in such plants. These were placed with the experiment at Medford, Mass., mentioned in Table 1, where birds took 69 per cent of the larvse from the cornstalks tied to stakes.

Of the several plants thus observed, common sunflower (. Helianthus annuus L.), Princesplume ( Polygonum orientate L.), Polygonum sp., and cocklebur ( Xanthium spp.) showed extensive feeding by birds of the same sort attributed to woodpeckers while no evidence of such feeding was noticed in Abutilon ( Abutilon theophrasti Medic.), pigweed ( Amaranthus retroflexus L.), rag- weed (. Ambrosia sp.,) beggar-ticks ( Bidens sp.), pot-marigold ( Calendula officinalis L.), aster ( Callistephus sp.), feather cocks- comb ( Celosia argentea L.), Cosmos bipinnatus Cav., Dahlia sp., barnyard grass ( Echinochloa crusgalli L.), Japanese millet ( Echi - nochloa sp.), Gladiolus sp., strawflower ( Helichrysum bracteatum

34

Psyche

[February

Andr.), geranium ( Pelargonium hortorum), lima bean ( Phaseolus lunatus, L.), and African marigold ( Tagetes erecta L.).

The field examinations of cornstalks showed much the same evidence as was obtained from the experimental work described in the preceding paragraphs. Numerous instances were found where birds had removed a high percentage of the larval content of cornstalks and such instances were found only in heavily

Fig. 1. Map of the area known to be infested by the European Corn Borer in New England in 1922. Circles show localities where experiments were placed in the fall of 1922; clear circles indicate that no feeding by birds was found, while circles having a cross in the center indicate localities where birds fed on larvae contained in the corn .stalks of the experiment.

A white line surrounds the area known to be infested up to July 1, 1919.

1 — Arlington, Mass.

2 — Medford, Mass.

3 — -Rockport, Mass.

4 — Newbury, Mass.

5— Wells, Me.

6 — Methuen, Mass.

7 — ’Farmington, N. H

8 — Tyngsboro, Mass.

9 — Bristol, N. H.

10 — Concord, Mass.

11 — -Worcester, Mass.

12 — Walpole, Mass.

13 — Taunton, Mass.

14 — Quincy, Mass.

15 — Scituate, Mass.

16— Wareham, Mass.

17 — Falmouth, Mass.

18 — Manomet, Mass.

19 — Harwich, Mass.

1925] Efficiency of Birds in Destroying Larvce of Corn Booer 35

infested areas and in localities that had been infested for several years. In sparsely infested localities or in areas where infesta- tion had been severe only recently, no extensive feeding was noticed although in some cases there were traces of feeding by birds. The following table shows the condition as found in certain heavily infested fields that were found to be attacked by birds.

TABLE II.

Extent of Feeding by Birds on Material Undisturbed by Man in the

Spring of 1923.

Locality	Date examined	Type of | corn	Condition of corn- stalks	Size of area	Estimated number of larvae taken by birds	Estimated per cent of larvae taken by birds
Watertown, Mass.	IV-2-23	Sweet	Standing
		1	and broken
			over	1 \ acres	186,480	92%
Watertown, Mass.	IV-6-23	U	standing and
			broken over	I acre	26,957	78%
Milton, Mass.	IV-13-23	“	Lying on
			soil	1 acre		30%
Marblehead,	V-3-23	u	standing and
Mass.			broken over	A acre	140,946	97%
Melrose, Mass.	V- 1 1-23	“	In piles
			partly burn-
			ed			80%

The figures mentioned in this table were obtained by counting the number of cornstalks in the several fields, estimat- ing the fall infestation of the stalks, and comparing this figure with the average infestation found on the date of the examina- tion. The spring infestation per cornstalk was obtained by averaging several series of counts made in different parts of each field. In all these instances both types of bird feeding shown in figure 1 were found, the feeding attributed to woodpeckers being extensive in standing stalks and the stalks lying on the ground being shredded by other birds.

In weeds instances were found of the removal of considerable numbers of larvae from cocklebur ( Xanthium sp.) and barnyard grass ( Echinochloa crusgalli L.). In one field several score of

36

Psyche

[February

dahlia plants were found from which birds had removed nearly all the larvae of the European corn borer that these plants had harbored.

Although birds (particularly the downy woodpecker) feed to a limited extent on the larvae of this insect in the fall and winter, most of the feeding is done in the spring, especially the shredding of stalks lying on the ground. The experiment placed in Medford, Mass., (table 1), showed on April 9, 1923, that birds had by that time taken 65 per cent of the larvae from the cornstalks. A similar experiment in the same locality was continued until June 27, 1923, a date when most of the insects had transformed to adults. At that time examination showed that birds had apparently taken 82 per cent of the insects from the cornstalks.

In the fall of 1923 a series of experiments much similar to those previously described was placed in the field for the purpose of obtaining information on the extent of feeding by birds on this insect. However, this time fifty stations were chosen instead of twenty as in the previous work. Each station was composed of four stakes to each of which was fastened five infested corn- stalks, the larval expectancy as obtained from stalk counts being- on an average of 15.5 larvae per stalk.

These stations were so selected as to cover the entire area infested by this insect in New England and were run out in lines as straight as possible from Arlington, Mass., as a center as fol- lows: first line comprising 12 stations in a northeasterly direction as far as Sebago, Me.; second line comprising 5 stations in a northerly direction as far as Farmington, N. H. ; third line com- prising 8 stations in a direction north by northwest as far as Bristol, N. H.; fourth line to the northwest as far as Gardner, Mass, comprising 3 stations; fifth line comprising 4 stations to the west as far as Worcester., Mass.; sixth line to the southeast as far as Touisset, Mass., composed of 4 stations; seventh line to the south as far as Mattapoisett, Mass., composed of 4 sta- tions; eighth line composed of 9 stations to the southeast and extending on Cape Cod as far as Provincetown, Mass. Fig. 2.

1925] Efficiency of Birds in Destroying Larvce of Corn Borer 37

Fig. 2. Map of the area known to be infested by the European Corn Borer in New England in 1923. Circles show localities where experiments were placed in the fall of 1923; — clear circles indicate that no feeding by birds was found, while circles having a cross in the center indicate localities where birds fed on larvae contained in the corn stalks of the experiment. A white line surrounds the area known to be infested up to July 1, 1919.

1 — Arlington, Mass.

2 — Medford, Mass.

3 — Saugus, Mass.

4 — Beverly, Mass.

5 — Marblehead, Mass.

6 — Rockport, Mass.

7 — Newbury, Mass.

8 — Hampton, N. H.

9 — Kittery, Me.

10— Wells, Me.

11 — Biddeford, Me.

12 — Scarboro, Me.

13 — Woburn, Mass.

14 — Andover, Mass.

15 — Methuen, Mass.

16 — Kingston, N. H.

17— Farmington, N. H.

18— Bedford, N. H.

19— Hillsboro, N. H.

20— Concord, N. H.

21— Franklin, N. H.

22 — -Bristol, N. H.

23 — Concord, Mass.

24 — -Westford, Mass.

25 — -Harvard, Mass.

26 — Leominster, Mass.

27 — Gardiner, Mass.

28 — Natick, Mass.

29 — Southboro, Mass.

30 — Worcester, Mass.

31 — Medfield, Mass.

32 — Foxboro, Mass.

33 — Attleboro, Mass.

34 — Touisset, Mass.

35 — -Needham, Mass.

36 — Bridgewater, Mass.

37 — Mattapoisett, Mass.

38 — Milton, Mass.

39 — W. Hingham, Mass.

40 — Cohasset, Mass.

41 — Marshfield, Mass.

42 — Kingston, Mass.

43 — Falmouth, Mass.

44 — Sandwich, Mass.

45 — Brewster, Mass.

46 — Wellfleet, Mass.

47 — Provincetown, Mass.

38

Psyche

[February

Examinations of the cornstalks of these experiments in the spring of 1924 showed that 16 of these stations exhibited feeding by birds to a noticeable extent, while the remainder showed either a mere trace of bird feeding or no evidence of such feeding at all. In Table 3 those stations that showed bird feeding are listed together with the number and percentage of larvae ap- parently removed from the stalks by birds.

The average winter mortality for the 38 stations listed in table no. 3 was 4 per cent.

The average per cent of larvae apparently taken by birds in the 16 stations that showed bird feeding (based on the average recovery per stalk in experiments not touched by birds) was 54.

The average per cent of larvae apparently taken by birds in the total number of 47 experiments (based on the average recovery per stalk in experiments not touched by birds) was 19.

Three stations were lost or destroyed, leaving 31 that showed either a mere trace of bird feeding or none at all. The infestation in the fall was on an average 310 larvae per station. Spring ex- aminations of the 31 stations that showed no important feeding by birds proved that the average infestation at that time was 14.7 larvae per stalk or 294 larvae per station, an apparent loss per station of 16 larvae or approximately 5 per cent of the ex- pected larvae. This loss is slight when it is remembered that in some stations birds apparently did take a few of the larvae, and that the cornstalks were necessarily handled several times and were transported for considerable distances, so that small pieces were sometimes broken off and lost. These figures, however, are offered to show that the findings as regards bird feeding, shown in Table no. 3, present a fair picture of the extent to which birds fed on this material.

These results show a considerable increase in the area in which birds fed extensively, (Fig. 2 and 3) over the results ob- tained in the spring of 1923. This may be due to the fact that the larger number of stations provided a much more accurate test of conditions and so gave a much better picture, or it may indicate a widening field over which birds have become aware of an existing food supply. There is also a possibility of the element of chance entering to the extent that if the right species of birds

1925] Efficiency of Birds in Destroying Larvce of Corn Borer 39

found the stations, feeding would result, whereas if the experi- ments remained undiscovered by birds able to take larvae from the cornstalks no evidence of feeding would be found. Whatever the reason, however, the fact of evidence of bird feeding on over- wintering larvae of this insect in cornstalks over a much more extended area during the spring of 1924 than was observed previously remains, and it is the writer’s belief that birds were a more important factor in reducing the numbers of the European corn borer in the spring of 1923 than in a corresponding period in 1922, and that in the spring of 1924 they were of greater im- portance than in the same period in 1923; in other words, that the importance of birds as a means of natural control has been increasing each spring for the last three years.

The figures showing the percentage of larvae taken by birds as shown in Table 3 represent the feeding up to the time that the experimental material was collected in the spring and so do not show the total amount of feeding that birds might have done had the material remained in the field a few weeks longer. As already mentioned, this same condition prevailed in the con- sideration of the experiments examined in the spring of 1923. It was necessary, however, to collect these experiments early in April because of a desire to examine the cornstalks before the larvae had moved from the exact locations in the stalks in which they rested at the time the material was set out the previous fall.

It is probable also that birds were unable to remove some of the larvae from the staked experiments because in tying corn- stalks to the stakes that part of the stalks lying next to the stakes was rendered inaccessible to the birds.

In the series of experiments examined in the spring of 1924, several of the experiments that showed no feeding by birds were located in areas that had been heavily infested by the insect for several years. Noticeable among those was the experiment located in Saugus, Mass, (table 3). This experiment was placed on a farm where corn had been severely infested each year since 1919. In this instance, however, little corn was grown in 1923 because of the heavy infestation previously experienced. A second instance of this condition was found in Marblehead,

40

Psyche

[February

TABLE NO. 3.

Extent of Feeding by Birds on Experimental Material in the Spring of 1924.

Experiments on which birds fed.

Locality	Date Placed 1923	Date Examined 1924	Total larval expectancy	Number of larvae recovered	Apparent num- 1 ber of larvae taken by birds	“ 1 Per cent of larvae taken by birds1	Per cent of larvae taken by birds.2
Arlington,	Mass.	11-14	4-13	310	125	185	‘ 59.6	57.4
Attleboro,	Mass.	11-28	4-11	310	81	229	73.8	72.4
Beverly,	Mass.	11-14	4-11	310	203	102	32.9	29.2
Brewster,	Mass.	11-24	4-14	310	81	229	73.8	72.4
Cohasset,	Mass.	11-20	4-11	310	201	109	35.1	31.6
Falmouth,	Mass.	11-23	4-12	310	118	192	61.9	59.8
Harvard,	Mass.	11-23	4-12	310	241	69	22.2	19.
Leominster,	Mass.	11.23	4-10	310	105	205	66.1	64.2
Marshfield,	Mass.	11-20	4-12	310	203	102	32.7	29.2
Medford,	Mass.	11-3	4-25	310	99	211	68.	66.3
Milton,	Mass.	11-23	4-12	310	131	179	57.7	55.4
Newbury,	Mass.	11-14	4-11	310	62	248	80.	78.5
Rockport,	Mass.	11-14	4-10	310	119	201	64.8	59.5
Southboro,	Mass.	11-27	4-10	310	77	233	75.1	73.4
Wellfleet.	Mass.	11-24	4-14	217	76	141	64.9	63.1
Woburn,	Mass.	11-14	4-12	310	237	73	23.5	19.3

Experiments on which birds did not feed

Andover,	Mass.	11-19	4-16	310	180
Bridgewater,	Mass.	11-27	4-10	310	262
Concord,	Mass.	11-16	4-12	310	256
Foxboro,	Mass.	11-28	4—1 1	310	262
Gardner,	Mass.	11-29	4-20	310	357
Hingham, Kingston,	Mass.	11-20	4-12	170	137	(partly destroyed)
Mass.	11-24	4-12	310	414
Marblehead,	Mass.	11-15	4-12	310	388
Mattapoisett,	Mass.	11-27	4-10	310	387
Medfield,	Mass.	11-24	4-11	310	318
Methuen,	Mass.	11-19	4-16	310	256
Natick,	Mass.	11-23	4-10	310	346
Needham,	Mass.	11-16	4-14	310	268
Provincetown,	Mass.	11-24	4-14	310	325
Sandwich,	Mass.	11-23	4-11	310	333
Saugus,	Mass.	11-14	4-11	310	303
Touisset,	Mass.	11-28	4-10	310	228
Worcester,	Mass.	11-27	4-10	310	228
Westford,	Mass.	11-15	4-12 -	310	357
Biddeford,	Me.	11-22 .	4-14	310	328
Kittery,	Me.	11-22	4-14	310	285
Scarboro,	Me.	11-22	4-14	310	295
Wells,	Me.	11-22	4-14	310	232
Bedford,	N. H.	11-21	4-17	310	281
Bristol,	N. H.	11-22	4-16	310	228	(partly destroyed)
Concord,	N. H.	11-21	4-16	263	203
Farmington,	N. H.	11-21	4-16	310	340
Franklin,	N. H.	11-21	4-16	310	181
Hampton,	N. H.	11-22	4-14	310	366
Hillsboro,	N. H.	11-21	4-16	310	200
Kingston,	N. H.	11-20	4-28	310	276

1 Based on the average larval contents of stalks in the fall.

2 Based on the average larval recovery at stations not attacked by birds.

1925] Efficiency of Birds in Destroying Larvce of Corn Borer 41

Mass., where birds did not feed on the larvae contained in the cornstalks of the experiment (table 3) although it was found in the spring of 1923 that they had removed a high per cent of the borers from a field of heavily infested cornstalks (table 2). From these instances it appears that birds might not be depended on to feed on corn borer larvae in cornstalks in the same locality each year. On the other hand, heavily infested localities are known, noticeably Medford, Mass., where birds have fed on overwintering larvae consistently in the springs of 1922, 1923 and 1924.

Throughout the infested area of Massachusetts there was very little corn standing in the field during the winter of 1923- 1924. The condition of all the experiments was the same, there- fore, in that practically no cornstalks other than the experiments were to be found by the birds and for this reason there was no influence brought to bear, as far as the extent of feeding was concerned, by proximity of the experiments to infested cornfields. Because of the general scarcity of standing corn during the winter of 1923-1924 it might appear that a condition of con- centrated feeding on the experimental material might result. It does not seem that any such phenomenon took place, however, because in the experiments examined in the spring of 1924, in no case was the bird feeding found to be as extensive as on the

TABLE NO. 4.

The Extent of Bird Feeding on Experimental Material at Medford, Mass., on different dates in the Spring of 1924.

Date Examined	Number of stalks	Number of expected larvae	Number of larvae recovered	Apparent number taken by birds	Apparent per cent taken by birds
IV-4-24	10	155	100	55	35-5
IV-11-24	10	155	73	82	53-
IV-17-24	10	155	62	93	60.
IV-24-24	10	155	108	47	30.3
V-2-24	10	155	50	I05	67.7
V-8-24	10	155	32	123	79-3
V- 1 5-24	10	155	18	137	88.4

42 Psyche [February

stalks of several cornfields examined in the spring of 1923 and listed in table 2.

When birds feed in the spring on larvae contained in corn- stalks that were piled up the previous fall and remained in such condition through the winter, an interesting phenomenon is frequently noticed. In the spring, larvae desert the wet corn- stalks in the lower parts of such a pile, migrating to the dry stalks above where conditions for transformation are much more favorable. It is on the larvae contained in these dry stalks on the top of the pile that birds such as grackles and blackbirds feed extensively so that as the spring advances it is frequently found that few larvae remain in the lower stalks of the pile because of the migration of the stalks above, and a few larvae re- main in the dry stocks on the top of the pile because birds have shredded the stalks and removed a high percentage of them. This condition has been found several times experimentally, and has been noticed in several localities in the field where cornstalks have passed the winter in piles.

Observations as to the extent of feeding by birds on larvae in cornstalks standing undisturbed in the field in the spring of 1924 were possible in only a few localities, mostly in very small lots of stalks because of a law in the state of Massachusetts compelling all persons to destroy standing corn in the fall of 1923. Those found, for the most part in small back-yard gardens, often showed evidence of extensive feeding by birds. Thus a small plot of about 1,500 hills of standing corn in Wakefield, Mass., showed that birds had removed a very high percentage of the larvae from the stalks. In various localities stubble and stalks lying on the ground were shredded and many larvae no doubt removed. As far as these observations were possible, they coincide with the results obtained in the experimental work already described.

The question has been asked whether birds know that infested cornstalks contain larvae or whether the feeding that they do is more in the nature of an accident. Beside the ex- periment at Medford, Mass., mentioned in table 3 from which birds are credited with removing 68 per cent of the larvae a like experiment was placed in the fall of 1923, similar in every respect

1925] Efficiency of Birds in Destroying Larvce of Corn Borer 43

except that the cornstalks showed no trace whatever of infes- tation by the European corn borer. Spring examinations showed that whereas birds had fed extensively on larvae in the infested stalks, there were only one or two shallow incisions that might be credited to birds in the stalks that were entirely free from infestation by this insect. These uninfested stalks were, there- fore, probably examined by the birds but the experiment showed that they were soon undeceived as to the fact that these stalks, although so much like the nearby infested ones in appearance, contained no larvae of the corn borer. Field observations have shown that birds also feed in the spring on the pupae of this insect, the nature of the feeding being similar in every respect to that described in the case of their feeding on larvae. In some cases portions of the pupae remain in the burrows of the insect as if the birds were not entirely pleased with the change that its food supply had undergone.

These studies have been confined mostly to the feeding by birds on the larvae of the European corn borer in the fall, winter and spring, at which time the insect is in overwintering quarters within the host plant, because no such extensive feeding has been observed during the growing season. Adults of the insect are active only at night, resting during the day on the under sides of leaves unless disturbed when their flight is short and low and is to the nearest cover from the seat of disturbance. The larvae for the greater part of their lives feed hidden within the food plants, leaving their burrows infrequently except during the warmer nights of midsummer. It is known, however, that birds readily take the adult of the insect although observations seem to indicate that up to the present time they have not taken larvae from growing plants in any noticeable numbers. On two occasions, during studies of the capabilities of flight of the moths, birds took the flying insects to such extent as to interfere seriously with the success of the experiments. It is probable, therefore, that the reason that they do not feed more extensively on the moths is not because of any distaste for them, but because of the inactivity of the moths during the hours of sunlight.

Since the larvae of the European corn borer in the over- wintering condition are nearly destitute of hairs and the skin is

44

Psyche

[February

rather thin and sparingly chitinized, they are ideal food for birds. Furthermore, at this time of the year the body of the larva is filled with so-called fat body so that the insect becomes food of the highest value with only a small portion of waste matter. The larvae in standing stalks, during times of heavy infestation, provide a plentiful supply of food for birds, such as woodpeckers, during the winter, provided the stalks are not completely covered with snow, at a time when other food is scarce. They also provide a source of food for migrating insectivorous birds in the spring, especially those arriving early before other insects become active and available in abundance. There is, therefore, every reason to hope that birds will take advantage of this food supply to the benefit not only of themselves, but also of man. This they may readily do, provided that they are able to locate the larvae in the stalks, as these experiments seem to indicate they are doing more and more, and if they are able to remove the larvae from the stalks after finding them, a feat of which not all insectivorous birds will be capable. However, there is ample evidence to indicate that their industry may help to hold the insect partially in check, or even so to reduce its numbers in some localities during the winter and spring that damage by the species may not be extensive enough to cause heavy loss to crops in such localities the following summer.

Summary.

Evidence of feeding by birds on larvae of the European corn borer has been found each spring for several years in New Eng- land. Such work was of two distinct types; the work of wood- peckers, particularly the downy woodpecker, which drills holes into standing stalks in order to reach the overwintering larvae of this insect; and work by grackles, blackbirds, starlings and other species which shred stalks that have fallen over and devour the hiding larvae.

For the most part such feeding by birds has been confined to localities that have been heavily infested for several years, the extent of such feeding having been found to vary greatly, but counts have shown that in some fields of sweet corn over

1925] Efficiency of Birds in Destroying Larvae of Corn Borer 45

90 per cent of the overwintering larvae have been removed from the stalks.

Experiments where infested cornstalks were placed in rep- resentative localities in the infested area of New England in the fall of 1922 and 1923 were examined the following spring in each case, before larvae had become active. These experiments were for the purpose of determining whether or not the habit of feeding on larvae of this insect was general. The experiments examined in the spring of 1923 showed that of 18 localities the cornstalks of which were recovered in good condition, birds had fed extensively on larvae contained in the cornstalks of 5 locali- ties, the proportion of larvae credited to such feeding ranging from 12 percent to 84 per cent, and averaging 61 per cent for these 5 localities and 17 per cent for the whole 18 experiments. Examined in the spring of 1924 showed that of 47 local- ities the cornstalks of which were recovered in good condition, birds had fed extensively on the larvae contained in the cornstalks of 16 localities, the extent of such feeding ranging from 19 percent to 78.5 per cent. The average proportion of larvae taken from these 16 localities was 54 per cent and the average of larvae taken from the whole number of 47 localities was 19 per cent. The feeding, which was mostly the work of woodpeckers, was found to be over a much more extended area in the spring of 1924 than in the spring of 1923.

Feeding by woodpeckers on the larvae of this insect over- wintering in sunflower, ( Helianthus annuus L.), Princesplume (Polygonum orientate L.), Polygonum sp., and cocklebur (Xan- thium spp.) have also been found in experiments and in barnyard grass (Echinochloa crusgalli L.) and cocklebur ( Xanthium spp.) in the field.

The importance of the feeding by birds on overwintering larvae of this insect at the present time, taking the infested area in New England as a whole, is not great, but in small areas the importance of their feeding must be considerable, since these small areas are often very heavily infested. The important point at present is that birds are finding the overwintering larvae of the corn borer, that they are feeding on them, and that such feeding seems to be on the increase.

46

Psyche

[February

Explanation of Plate I.

At left. The appearance of corn stalks after woodpeckers have fed on the larvae of the European corn borer that the stalks harbored. The section to the right shows the holes made by these birds in reaching the larvae. The two sections to the left are of one corn stalk split open to show the condition of the center of the stalk, all the larvae having been removed by these birds.

At right. Corn stalks infested by the European corn borer showing the appearance of the stalks after birds such as blackbirds and grackles have broken them open and devoured the larvae that they contained.

PSYCHE 1925 VOL, XXXI'. PLATE I.

BARBER— EUROPEAN CORN BORER.

1925]

A New Species of the Genus Gaurax

47

A NEW SPECIES OF THE GENUS GAURAX.

By Charles W. Johnson.

Boston Society of Natural History.

Gaurax dorri sp. nov.

Upper part of front, vertex, and occiput dull black, with short black hairs, lower part of front and orbits yellow, face whitish, oral margin above black, cheeks with a narrow black line which extends as a faint brownish line to the base of the antennae, the latter yellow, the margin of the third joint and arista dark brown. Thorax and pleura black, shining, with two small yellow spots near the base of the scutellum, the latter yellow, the disc slightly darkened. Abdomen black, shining, the middle of the first segment and venter yellow, in the female the first segment is entirely yellow. Front coxae black, middle and hind coxae yellow, femora black, with bases and apices yellow, front and middle tibiae and tarsi yellow, posterior tibiae black, first and second joints of the posterior tarsi yellow, the others black. Halteres yellow. Wings hyaline. Length 2 mm.

Holotype and allotype, Great Pond, Mt. Desert, Maine, June 27, 1922 (C. W. Johnson). In the collection of the Boston Society of Natural History. The species is dedicated to Mr. George B. Dorr, Director of the Lafayette National Park, who has greatly aided in the work on the insect fauna of Mt. Desert.

48

Psyche

[February

NEW NEOTROPICAL THYSANOPTERA COLLECTED BY C. B. WILLIAMS.

By J. Douglas Hood,

University of Rochester.

The present paper brings forward several new genera and a number of new species collected by Mr. C. B. Williams during his residence in Trinidad from 1916 to 1919 and also during a short trip to the Lesser Antilles in March and April 1915.

It was originally intended by Mr. Williams that the material be worked up by us jointly; but my own delay in getting at the task until he had been appointed to the distant post of Ento- mologist to the Egyptian Ministry of Agriculture has made such co-operative work seem inadvisable. During his various visits to the United States the systematic position of many of the genera and species was worked out by Mr. Williams and ten- tative manuscript names assigned. These names have been retained except in the case of several of the more interesting and conspicuous forms which, I think, are better named after their collector in recognition of his exceedingly valuable work on this group of insects.

The following descriptions are preliminary to other papers in press or in preparation, and are published at this time in order that the worth of this splendid collection — certainly the largest ever brought out of tropical America — may not be lessened by the otherwise inevitable loss of priority in many of the names.

The holotypes, allotypes, and a portion of the paratypes have most generously been placed by Mr. Williams in my col- lection.

Stomatothrips septenarius sp. nov.

Female (macropterous) . — Length about 1.4 mm. Color testaceous, with head, prothorax, two pterothoracic transverse bands (one near middle and one at base), abdominal segments 3-9, all tibiae and tarsi, and antennal segments 4-9, blackish brown; antennal segments 1-3 yellowish white, 4 paler basally; fore wings pale gray-brown, with two white cross bands, one narrow and at basal sixth, the other wider and at apical seventh,

1925]

New N eotropical Thysanoptera

49

intermediate brown area somewhat paler at middle; hind wings light gray with two pale bands corresponding in position with those on fore wings. Maxillary palpi sei;en-segmented, the basal segment large; labial palpi five-segmented, the basal segment short. Fore wings somewhat expanded apically, broadest at apical sixth, where they are about 1.6 times as wide as at basal fourth.

Trinidad and St. Thomas; C. B. Williams; taken by sweep- ing.

The seven-segmented maxillary palpi, the dark brown fourth antennal segment and the narrower fore wings distinguish this species at once from S. flavus Hood, the type of the genus, to which it bears a close superficial resemblance.

Frankliniella parvula sp. nov.

Female (macropterous). — Length about 1.1 mm. Color light brown, apex of abdomen darkest; femora usually brown, much paler at apex, tibiae and tarsi pale yellowish gray; antennae with segments 1, 2 and 478 brown, 2 darkest, 3 light yellowish gray, 4 and 5 paler basally; wings pale brownish, narrowly lighter along median line; ocellar pigment dull orange. Head about 1.33 times as wide as long. Segment 2 of antennae slightly thickened on dorsum at apex and bearing 2 rather prominent dark bristles; length of antennal segments in microns: 1, 24; 2, 36; 3, 75; 4, 51; 5, 39; 6, 53; 7, 10; 8, 17.

Male (macropterous). — Pale yellow, with a gray-brown blotch at middle of abdominal tergites 2-8; wings almost clear; antennae nearly white, segments 4 and 5 tipped with gray, 6-8 gray.

Trinidad, Grenada, Panama, and Costa Rica; C. B. Wil- liams; on cacao, in various flowers, etc.

Easily recognized by the long third antennal segment. The prolongation of the second segment can be seen to advantage only when viewed from the side.

50

Psyche

[February

Frankliniella nigricauda sp. nov.

Female (macropterous). — Length about 1.4 mm. Color

bright orange-yellow, head paler, abdomen shading from yellow at base to dark blackish brown or black in apical segments.

Trinidad; C. B. Williams; swept.

The coloration is unique in the genus.

Anaphothrips bicinctus sp. nov.

Female (macropterous). — Length about 1.0 mm. Color dark brown (tip of abdomen darkest), with prothorax, ab- dominal segments 3-6, legs and segments 3 and 4 of antennae, abruptly pale yellow; segment 5 of antennae gray, 6-8 dark brown; wings light gray, with a dark, brownish band across fore wings from basal fifth to near middle.

Trinidad; C. B. Williams; swept.

Very close to A. alternans (Bagnall), but separable by the smaller size and the coloration of the abdomen.

Astrothrips angulatus sp. nov.

Female (macropterous). — Length about 1.0 mm. Dorsal surface deeply reticulate. Color yellowish brown, with base of abdomen and sides of pterothorax darker; abdominal segments 8 and 9 with a distal gray band; antennae yellow, with segments 1 and 2 darker and apices of segments 4, 5 and 6, and all of seg- ment 7, brownish; fore legs yellow, the femora and tibiae brown- ish at sides, especially near middle; middle and hind femora and middle tibiae brown, yellow at either end; middle and hind tarsi and hind tibiae yellow, the last clouded with brown in apical third; fore wings brown, with three transverse, narrow, white bands, situated at basal fifth, apical two-fifths, and ex- treme apex, respectively. Vertex slightly elevated but not at all produced or overhanging. Antennae seven-segmented; seg- ment 1 short, subcylindrical, about as broad as long; 2 the broadest in entire antenna, goblet-shaped, pedicellate; 3 slender, vasiform, pedicellate, about four times as long as wide; 4 about

1925]

New N eotropical Thysanoptera

51

0.8 as long as 3 and somewhat stouter; 5 about 0.7 as long as 4 and of the same width, oval, pedicellate; 6 equal in length to 5 but more slender, of the same form but inverted, the pedicel being apical and delimited by an oblique transverse line; 7 shorter, very slender; sense cones simple. Pro thorax three- fourths as long as head. Abdomen strongly and sharply cons- tricted beyond base of segment 2, which is the longest in entire abdomen; 10 tubular, about 0.7 as wide at base as long, divided above by a longitudinal suture.

Grenada, Guadeloupe and Trinidad; C. B. Williams; on cacao, an unidentified creeper and fustic ( Chlorophora tinctoria).

Readily known by the seven-segmented antennae with simple sense cones, the non-produced vertex, and the short prothorax, which is angulate at the sides.

Astrothrips constrictus sp. nov_

Female (macropterous).— Length about 1.1 mm. Dorsal surface deeply reticulate. Color yellowish brown, paler at middle of pterothorax and in last three abdominal segments; segment 10 of abdomen nearly black in apical third; antennae yellow, with segments 1 and 2 darkened with brown, and apical three-fifths of segment 5 and all of segment 6 unevenly darkened with blackish brown; femora brown; tibiae brown, narrowly yellow at either end; tarsi yellow; fore wings brown at base (scale darker) and with brown blotches of varying lengths scat- tered along veins, these blotches forming more or less distinct transverse bands at basal third and apical fourth; ring vein at extreme tip of wing dark brown. Vertex produced, overhanging the insertion of antennae. Antennae six-segmented; segment 1 short, subcylindrical, about as broad as long; 2 broadest in entire antenna, goblet-shaped, pedicellate; 3 slender, vasiform, pedicellate, fully five times as long as wide; 4 somewhat more than half as long as 3 and slightly stouter, subfusiform, pedi- cellate, with an oblique, transverse line beyond middle; 6 about one-third as long as 5, slender; sense cones slender, those on segments 3 and 4 bifurcate beyond base. Abdomen strongly and sharply constricted beyond base of segment 2, which is the

52

Psyche

[February

longest in the entire abdomen; 10 strongly constricted at basal fourth, widened beyond, and narrowed again at apex, divided above by a longitudinal suture.

Trinidad; C. B. Williams; on Lantana sp.

Recognizable at once by the basal constriction of the tenth abdominal segment and the six-segmented antennae with their Y-shaped sense cones.

Goremothrips gen. nov.

(K opeya, a broom; Opajy, a wood worm.)

Body and all appendages very slender, and all bristles of monstrous size, hooked and pinnatifid apically. Head with two pairs of such bristles, one interocellar and the other postocellar; prothorax with two at each of its four angles. Wings very narrow; median vein of fore wing fused with costa, the strengthened anterior margin armed with about fifteen of the usual tremendous bristles disposed in two series, one inclined forward and the other backward; fringing hairs weak and sparse. Segment 9 of abdomen with two pairs of enlarged bristles, segment 10 with one pair.

Genotype : Coremothrips pallidus sp. nov.

A striking genus, of undoubted affinities with Scolothripst but far more extreme than even that bizarre form in the de- velopment of the bristles of the body and fore wings.

Coremothrips pallidus sp. nov.

Female (macropterous) .^Length about 0.9 mm. Color uniform, pale yellowish white.

Trinidad, Panama, and St. Vincent; C. B. Williams; on leaves of cacao and avocado pear.

Plesiothrips octarthrus sp. nov.

Female (macropterous). — Length about 0.9 mm. Head, prothorax and abdomen brown, the latter darker posteriorly; pterothorax pale brownish yellow; legs pale yellow; antennae

1925]

New N eotropical Thysanoptera

53

uniform dark brown, somewhat darker than head, except seg- ment 3 which is pale brown, with pedicel and apex pale gray; wings brown, the fore pair without pale band at base, but with an indistinct, narrow, median, pale streak. Antennae 8-seg- mented; segment 3 decidedly less than twice as long as wide. Segment 10 of abdomen divided above.

Trinidad; C. B. Williams; no further data.

An undoubted Plesiothrips, remarkable for the distinctly eight-segmented antennae.

Plesiothrips amblycauda sp. nov.

Female (macropterous). — Length about 1.1 mm. Color brown, head darkest, prothorax with bright crimson subhypo- dermal pigmentation; femora brown, the fore pair yellow at apex; tibiae gray, fading to pale yellow apically; tarsi pale yellow; antennae dark brown, about concolorous with head, except segment 3, which has the pedicel and apex pale gray; wings dark brown, without pale band at base. Segment 3 of antennae more than twice as long as wide. Segment 10 of ab- domen not divided above. Ovipositor extremely minute, ves- tigial, entirely functionless; tip of abdomen rounded, rather than conical.

Trinidad; C. B. Williams; on leaf of Arum lily.

The long third antennal segment and the undivided tenth abdominal segment, together with the dark coloration, make this a very distinct species. It is a true Plesiothrips, with an- tennae presumably seven-segmented, though the extreme tips of both are missing in the unique type.

Merothrips cognatus sp. nov.

Female (macropterous). — Length about 1.0 mm. Very close indeed to M. fusciceps Hood and Williams, but with head smaller (width 0.105 mm.) and eyes larger, these prolonged on ventral surface to a point directly beneath posterior dorsal margin of head, and about, equal in width to their interval; tooth on fore tibia minute.

54

Psyche

[February

Antennal segments: 12345678

Length (/*) 18 33 36 32 20 25 27 33

Width 0) 29 27 22 23 17 17 16 13

Trinidad; C. B. Williams; on dead branch of Lagerstroemia infested with bromeliads.

This species is separable from williamsi Priesner, described from Paraguay, by the much shorter and stouter antennal segments, particularly the fourth, which in williamsi is nearly twice as long as wide.

Adraneothrips gen. nov.

(aspavrjs, feeble; Opi\p, a wood worm.)

Allied to Haplothrips, but wings narrowed and parallel-sided beyond base, not constricted at middle, sparsely fringed. Body weakly chitinized. Eyes often prolonged on ventral surface of head. Terminal segment of antenna long and slender, not closely united to 7.

Genotype : Haplothrips (?) tibialis Hood. The species des- cribed by Hood and Williams as Haplothrips (?) bellus also belongs here, together with the five new species described below.

Adraneothrips simulator sp. nov.

Female (macropterous). — Length about 1.1 mm. Bicolo- rous; thorax and all of abdomen except tube pale yellow, with a band of bright crimson subhypodermal pigmentation along each side, broadest in pterothorax and abdominal segments 4 and 5; head and tube light brown, segments 8 and 9 of abdomen lightly shaded with brown; antennae with most of segment 3 and base of 4 pale; tibiae pale yellow or grayish white; femora shaded with brown. Eyes prolonged on ventral surface of head beyond their posterior dorsal margins; anterior marginal bristles of prothorax long and knobbed.

Male (macropterous). — Smaller and more slender than female.

1925]

New N eotropical Thysanoptera

55

Trinidad; C. B. Williams; “along midrib on under surface of leaves of undetermined plant.”

Like A. alternatus in color, but with the eyes distinctly prolonged on ventral surface of head.

Adraneothrips abdominalis sp. nov.

Female (macropterous) . — Length about 1.1 mm. Bico- lorous; head, thorax, and abdominal segments 1 and 7-10 brown, head darkest, no subhypodermal pigmentation; antennse with most of segment 3 and bases of 4-6 pale; segments 2-6 of abdomen pale yellow, or 6 lightly brownish; tibiae pale yel- low or grayish white; femora not, or only slightly, darker. Eyes prolonged on ventral surface of head beyond their posterior dorsal margins. Anterior marginal bristles of prothorax long and knobbed.

St. Thomas, West Indies; C. B. Williams; grass.

Easily separable by the abdominal coloration and prolonged

eyes.

Adraneothrips alternatus sp. nov.

Female (macropterous). — Length about 1.1 mm. Bico- lorous; head, pterothorax and abdominal segments 4 and 5 and 8-10 light brown, much darker than intervening portions and with bright crimson subhypodermal pigmentation; prothorax pale yellow, with a narrow band of crimson pigmentation along sides; legs pale; antennae with bases of segments 3-6 pale. Head 1.1 times as long as wide; eyes not prolonged on ventral surface of head, about equal in width to their interval and nearly as long as their distance from posterior margin of head.

Male (macropterous). — Smaller and more slender than female, and usually paler in coloration; segment 8 of abdomen usually much paler than 9, often concolorous with paler portions of abdomen.

Trinidad and Panama; C. B. Williams; common on dead banana leaves, on upper side along midrib.

56

Psyche

[February

Adraneothrips fuscicollis sp. nov

Female .(macropterous). — Length about 1.1 mm. Bico- locous; head, prothorax, pterothorax and abdominal segments 4 and 5 and 8-10 brown, much darker than the intervening portions and with bright crimson subhypodermal pigmentation; legs pale: antennae with most of segment 3 and bases of 4-6 pale. Head nearly as wide as long; eyes not prolonged on ventral surface of head, three-fourths as long as their distance from posterior margin of head, about two-thirds as wide as their interval.

St. Lucia, British West Indies; C. B. Williams; from mis- cellaneous bushes.

Adraneothrips uniformis sp. nov.

Female (macropterous). — Length about 1.1 mm. Uni- colorous; pale brown, with bright crimson subhypodermal pig- mentation at front and along sides of head, throughout thorax, along sides of abdomen, and in last two abdominal segments; antennae nearly uniform brown, segment 3 paler basally; legs brown, with trochanters, tarsi, and both ends of all tibiae, pale. Head 1.09 times as long as wide; eyes not prolonged on ventral surface of head, about equal in width to their interval and only slightly more than one half as long as their distance from posterior margin of head.

Trinidad; C. B. Williams; on dead banana leaves with A. alternatus.

Zygothrips speciosus sp. nov.

Male (macropterous). — Length about 1.0 mm. Color dark brown, with abdominal segments 1-6, all tibiae and tarsi, and antennal segments 3-5, clear pale yellow, almost white; segment 2 of antennae pale apically, 6 pale in basal half; wings nearly colorless. Postocular bristles nearly pointed. Prothorax with distinct bristles at posterior angles only, outer pair broadly dilated at tip, inner pair nearly pointed.

1925]

New N eotropical Thysanoptera

57

St. Croix, Dutch (now American) West Indies; C. B. Wil- liams; from grass.

The only species of the genus so colored.

Hindsiana rhopalocera sp. nov.

Female (macropterous). — Length about 1.1 mm. Color yellow; head decidedly darkened with brown, particularly an- teriorly; prothorax and sides of pterothorax lightly brownish; abdomen shading to pale grayish brown in segments 8 and 9; tube nearly black except for a narrow, pale, brown band at base and one of equal width at apex; legs uniform pale yellow; antennae yellow in segments 1-3, uniform dark blackish brown beyond, segment 1 shaded with brownish basally; wings faintly yellowish at base, nearly clear beyond. Head about 1.3 times as long as wide, cheeks parallel; eyes less than one-fourth as long as head, about two-thirds as wide as their interval; postocular bristles long, broadly dilated apically; antennae with segments 7 and 8 compactly united to form a heavy club which is twice the length of segment 6, segment 8 less than half as long as 7, which is distinctly the longest in entire antenna. Prothorax with all bristles present, subequal to or shorter than postoculars and similarly dilated; median thickening distinct. Wings of fore pair without accessory hairs; outer subbasal bristle short and pointed, others dilated, about as long as those at anterior angles of prothorax. Tube about 0.6 as long as head, less than twice as long as basal width, which is less than twice the apical; sides somewhat concave; terminal bristles more than twice the length of tube.

Guadeloupe, French West Indies; C. B. Williams; from an undetermined plant.

The structure of the last two antennal segments distinguish this species from its congeners.

Trichothrips calcaratus sp. nov.

Female (apterous). — Length about 2.3 mm. Color brownish yellow, with prothorax, pterothorax and basal three-fourths of

58

Psyche

[February

tube darkened with brown; extreme tip of tube light gray; abdominal segments 3-8 with a dark brown transverse band on upper surface near base; antennal segments 1 and 2, and basal half of segment 3, yellow and concolorous with head, remainder of antenna dark blackish brown except for the yellowish pedicels of segments 4 and 5; hypodermal pigmentation orange, opaque. Head with two, more or less evident, setose lateral tubercles just behind eyes. Fore tibia produced at lower, inner surface of apex to form a flattened, obtuse spur which is about one-third as long as width of tibia; fore tarsus with a large, straight, acute tooth. Tube 0.75 as long as head.

Male (apterous). — Much like female, but with the two lateral tubercles just behind eyes well-developed and usually very prominent. Fore tibia with an acute, additional projection on inner surface of apex, directly above the flattened, obtuse spur; fore tarsus with the tooth longer than width of tarsus.

Trinidad; C. B. Williams; from dead branch of Lager s- troemia infested with Bromeliads, and from bamboo.

The form of the head is suggestive of T. flavicauda Morgan. It is the only species of the genus with strongly armed fore tibiae.

Eurythrips collaris sp. nov.

Male (macropterous). — Length about 1.1 mm. Color blackish brown with purple subhypodermal pigmentation; legs paler, mid and hind tibiae and all tarsi, pale yellow, fore tibiae shaded with brown. Antennae with basal half of segment 3, basal third of 4, and basal fourth of 5 pale yellow. Vertex sharply conical. Antennae long and slender, all segments elon- gated. Mouth cone long, fully attaining posterior margin of prosternum. Bristles long, knobbed; prothorax with anterior angulars wanting, all others present, anterior marginals shortest. Fore tarsus unarmed. Wings slender, sparsely fringed, 2 or 3 accessory bristles. Tube very much shorter than head.

Trinidad; C. B. Williams; from dead tree in forest.

The only species of the genus without long bristles at the anterior angles of the prothorax.

1925]

New N eotropical Thysanoptera

59

Eurythrips gracilicornis sp. nov.

Female (macropterous). — Length about 1.1 mm. General color yellow, with pterothorax and tube dark brown; head brownish yellow, shading to dark brown at sides; legs nearly uniform yellow, femora lightly shaded with brown; abdomen brownish at sides and apically; antennae brown (excepting segment 3 which is clear yellow), segments 1 and 2 darkest, 1 paler at base, 2 paler at apex, 4-6 paler at base and apex. All segments of antennae, excepting 1 and 2, very long and slender for the genus, 4 and 5 three times as long as wide. Mouth cone very short, broadly rounded, hardly attaining middle of pros- ternum. Bristles long and dilated at apex; prothorax with anterior angulars well developed and anterior marginals wanting. Fore tarsus with a long, hooked tooth. Wings brown, fore pair without accessory bristles. Tube very much shorter than head.

Trinidad; C. B. Williams; swept.

Distinguishable by the short mouth cone and tube, and the slender antennae.

Lissothrips breviceps sp. nov.

Female (brachypterous). — Length about 0.9 mm. Color dark brown with red subhypodermal pigmentation; segments 1, 2 and 4-8 of antennae nearly concolorous with body, 2. slightly paler, 3 light gray, much the palest in entire antenna; legs dark, tarsi slightly paler. Head hardly 0.8 as long as wide; segment 3 of antennae conspicuously small and weak. Prothorax with all bristles present and broadly dilated at tip. Fore tarsus unarmed. Tube about 0 . 8 as long as head, 1 . 4 times as long as basal width, which is distinctly more than twice the apical.

Male (brachypterous). — Essentially like female but smaller.

Trinidad; C. B. Williams; from dead branches.

The antennal coloration, short head, and short, broad tube are distinctive.

60

Psyche

[February

Lissothrips pallipes sp. nov.

Female (apterous). — Length about 1.1 mm. Color shading from bright yellow in anterior half of head, through yellowish brown at sides and posterior part of head and in thorax, to nearly black in abdominal segments 5-10, tube brown in apical two- fifths; legs and antennal segments 1 and 2 bright yellow; anten- nae shading from yellowish gray in segment 3 to dark brown at tip. Head about 1.1 times as long as wide. Prothorax with anterior marginal bristles wanting, anterior angulars short, mid- laterals shorter than postoculars and shorter than the two pairs at the posterior angles, all bristles pointed. Fore tarsus unarmed. Tube less than 0.8 as long as head and 1.4 times as long as basal width, which is distinctly less than twice the apical.

Trinidad; C. B. Williams; from faggots and branches.

The coloration, long head and pointed bristles readily distinguish this species.

Williamsiella gen. nov.

Head and thorax very small; abdomen exceedingly large and broad. Eyes small. Antennae 7-segmented; segment 3 very small, shorter and narrower than any of the following segments. Mouth cone long, broadly rounded at tip, fore tarsus unarmed.

Genotype : Williamsiella bicoloripes sp. nov.

Evidently a derivative of Lissothrips. Separable by the 7-segmented antennae.

Williamsiella bicoloripes sp. nov.

Female (apterous). — Length about 0.9 mm. Color nearly u&iforbri dark blackish brown, tube paler apically; segments 1 and 2 of antennae and all femora decidedly paler, the femora somewhat darker in basal half or more; all tibiae and tarsi and antennal segments 3-7 (except the pale pedicel of 3) dark black- ish brown. Head wider than long; postocular bristles pointed, equal in length to head. Prothorax decidedly longer than head,

1925]

New N eotropical Thysanoptera

61

with prominent pointed bristles at posterior angles only, the outer pair much longer than prothorax. Fore tarsus unarmed. Tube more than half as long as head, nearly as wide at base as long (!).

Male (apterous). — Like female in all essential respects, but smaller.

Trinidad; C. B. Williams; on faggots.

Easily known by the seven-segmented antennse, long- bristles, exceedingly short and broad tube, and the coloration of the legs and antennse.

Plectrothrips impatiens sp. nov.

Female (macropterous). — Length about 1 .7 mm. Prothorax with midlateral and anterior marginal bristles longer than post- oculars. Antennal segments 3-5 each with five (!) sense cones, 6 with three.

Trinidad; C. B. Williams; on cacao leaf.

The only species of the genus with well-developed midlateral and anterior angular bristles on the prothorax, and the only one with more than three sense cones on any antennal segment.

Pristothrips gen. nov.

(777KST1?, a saw; OpL\f/, a wood worm.)

Head much longer than wide, cheeks with one or two strong- bristles at basal third or two-fifths; eyes large, much wider than their interval, reniform as seen from above and very closely facetted; intermediate antennal segments moderately elongated, clavate, sense cones not unusually long; mouth cone slender, long and pointed. Fore femora enlarged in both sexes, always with a large subapical tooth on inner surface, sometimes with a row of small teeth basally; fore tibise with several (3-8 in the two known species) strong teeth on inner surface; fore tarsus with one long tooth. Wings broad, narrower apically, not cons- tricted at middle, without venation. Tube' shorter than head.

Genotype : Pristothrips aaptus sp. nov.

62

Psyche

[February

Strongly suggestive of A canthothrips and its allies in the armature of the apex of the fore femur and the general details of body structure, but differing from them all in having the fore tibiae armed with several teeth. In several respects it approaches Machatothrips , in which, however,' the fore tibiae are unarmed. Ischyrothrips and Macrothrips lack the subapical femoral tooth. Eupathithrips has the prominent femoral tooth, but the cheeks are set with large spiniferous tubercles, and the urn-shaped intermediate antennal segments are provided with exceedingly long, slender, sense cones.

Pristothrips albipunctatus sp. nov.

Male (macropterous). — Length 3.0 mm. Color dark brown, abdominal segments 3-7 each with a pair of small, snow-white, lateral spots; intermediate antennal segments yellow, irregularly mottled with brown, 4-6 with a touch of dark blackish brown at extreme base; femora brown; fore tibiae yellow, mottled with brown, mid and hind tibiae brown, distinctly yellowish at either end; tarsi yellow. Head about 1.8 times as long as wide, with two pairs of long, stout bristles in basal two-fifths of cheeks, the apical bristle longer. Prothorax about 0.6 as long as head; bristles at anterior angles exceedingly slender, nearly as long as prothorax, blunt; bristles at posterior angles two-thirds as long, much stouter, blunt; all other bristles wanting. Fore wings with about 30 accessory hairs; two outer subbasal bristles long and pointed, the other shorter and blunt. Fore femur with two subapical teeth on inner surface, the basal one largest and tri- angular; fore tibia with a row of three large, stout teeth along inner surface, the basal one paired with a fourth one standing alone in another row; fore tarsus with a very long, stout, straight tooth. Tube two-thirds as long as head, 2.5 times as long as basal width, which is fully twice the apical.

Trinidad; C. B. Williams; on bark and faggots.

Pristothrips aaptus sp. nov.

Female (macropterous). — Length 3.2 mm. Color dark brown or black, with intermediate antennal segments irregularly slightly

1925]

New N eotropical Thysanoptera

63

paler; legs concolorous with body. Head about 1.9 times as long as wide, as broad across eyes as at middle, narrowed to base, with a pair of long, stout bristles at basal two-fifths of cheeks. Prothorax about 0.4 as long as head, anterior margin deeply roundly emarginate, posterior margin perfectly straight; bristles at anterior angles dilated apically, very short, shorter than those at posterior two-fifths of cheeks; one pair of bristles at posterior angles, four times as long as those at anterior angles and dilated apically: all other bristles wanting. Fore wings with about 35 accessory hairs on posterior margin; outer subbasal bristle long and pointed, others shorter, dilated at tip. Fore femora with two subapical teeth on inner surface, the basal one largest and triangular, and followed by a row of about eight smaller, spiniferous tubercles, of which the basal is decidedly the largest; fore tibiae with a row of 7 or 8 strong teeth along inner surface, the apical one longest, the others successively shorter; fore tarsus with a longer, stout, straight tooth. Tube three-fourths as long as head, fully 2.5 times as long as basal width, which is more than twice the apical.

Trinidad; C. B. Williams; on cacao.

The more distal of the two subapical teeth on the inner surface of the fore femur could easily be overlooked in a dark specimen, or one in which the femur is not in just the right position.

Neurothrips williams! sp. nov.

Female (macropterous). — Length about 1.9 mm. Very

close to N. magnafemoralis (Hinds) in general structure and in color, but without the femoral tooth of that species. Tube one- half as long as head, uniform blackish brown; segment 9 of abdomen pale brown.

Panama; C. B. Williams; on faggots.

Macrophthalmothrips williams! sp. nov.

Female (macropterous).— -Length about 1.4 mm. Color

yellowish white, with the space between eyes, and abdominal

64

Psyche

[February

segments 6-10, abruptly nearly black; antennae nearly white, with segment 2 at sides, all of 5 except pedicel, and all of 7-8 dark brown; legs nearly white, mid and hind tibiae with a narrow ring of gray near middle; cup of tarsi dark; wings colorless; eyes bright red, a dash of brown behind each; mesothorax lightly marked with brown at sides; metathorax with a brown spot at middle of metascutum and one at each posterior angle; abdominal segments 2 and 3 each with two pairs of minute brown dots, one pair latero-dorsal, the other latero-ventral ; segment 4 with the latero-ventral pair only; segments 6-10 with bright red subhypodermal pigmentation,

Trinidad; C. B. Williams; on dead branch of Lagerstroemia.

A very beautiful species, as strikingly colored, perhaps, as any in the family. Named after Mr. C. B. Williams, its dis- coverer.

Cryptothrips gradatus sp. nov

Female (brachypterous). — Length about 1.6 mm. Color yellow; abdomen shading from brownish yellow in segment 2 to nearly black in basal three-fifths of tube, apex of tube pale; head in front of eyes, and first two antennal segments brownish yellow; segment 3 of antennae brown, pedicel yellowish; 4-8 dark blackish brown; legs uniform yellow. Head 1 . 5 times as long as wide, narrowed behind eyes and at base; eyes small, composed of a few large, separated facets; ocelli wanting; postocular and post- ocellar bristles subequal, knobbed; dorsum of head with a pair of pointed bristles behind postoculars. Segment 8 of antennae conical, closely united to 7. Fore tarsi unarmed. Prothoracic bristles all present, knobbed, those at posterior angles longer than postoculars, others subequal to postoculars. Tube about 0 . 6 as long as head, less than twice as long as width at base, which is somewhat more than twice the apical.

Tobago, British West Indies; C. B. Williams; from grass.

A true Cryptothrips, allied to C. icarus Uzel, but readily known by the color, and the form of the last antennal segment.

1925]

New Neotropical Thysanoptera

65

Cryptothrips acuticornis sp. nov.

Female , forma macroptera. — Length about 1 . 6 mm. Color nearly uniform dark brown, head somewhat darker than rest of body; tarsi, fore tibiae and articulations of legs paler, as is also the apical portion of segment 2 of the antennae and all of segment 3 except the brown apex and a barely perceptible shading at basal two-fifths; segments 4 and 8 of antennae uniform dark brown. Head very slightly longer than wide, roundly converging from eyes to base; vertex tumid; eyes more than one-third as long as head, slightly more than one-half as wide as their interval; ocelli small, widely separated; postocular bristles blunt but not dilated at tip; postocellars and mid-dorsal pair minute. Segment 8 of antennae long and slender, four times as long as greatest width, not pedicellate. Fore tarsi unarmed. Wings not narrowed at middle, brownish at extreme base; fore pair with 6 or 7 ac- cessory hairs and with only two subbasal bristles, which are blunt at apex. Prothoracic bristles all present, hardly pointed, the two pairs at posterior angles longest and subequal to post- oculars. Tube 0.9 as long as head, more than twice as long as width at base, which is somewhat less than twice the apicftl #

Female , forma brachyptera. — Almost indistinguishable from the macropterous form save for the short wings.

Male (brachypterous) . — Decidedly paler than female, the tube and head dark brown and the intervening portion brownish yellow, darker posteriorly. Fore tarsus armed.

St. Croix and Barbados, West Indies; C. B. Williams; from grass.

A true Cryptothrips. The structure of the terminal antennal segment is distinctive.

Cryptothrips connaticornis sp. nov.

Male (macropterous). — 'Length about 1.4 mm. Color bright yellow, with anterior part of head and all of prothorax brown; mesothorax brownish at sides and along anterior margin; abdomen lightly shaded with brown in basal three or four seg-

66

Psyche

[February

ments; tube yellow, tipped with gray; subhypodermal pigmen- tation maroon red, disposed in head, thorax and abdomen wherever they are shaded with brown and also along sides of abdomen; antennae bright yellow in segments 1 and 2, segment 3 grayish at tip, 4 shading from yellowish gray at base to brownish gray apically, remaining segments successively darker, 5 and 6 somewhat lightened basallv, 7-8 dark blackish brown; mid and hind femora yellowish gray, remainder of legs clear yellow. Body bristles expanded but not divided at apex.

Trinidad; C. B. Williams; from faggots.

Structurally almost indistinguishable from the North American Cryptothrips junctus Hood. The figures of that species given in the Canadian Entomologist, Yol. XLIV, p. 140, 1912, would serve almost equally well for this. The coloration, how- ever, is utterly different.

Female (macropterous). — Length about 2.1 mm. Color brown, head darker than thorax, abdomen shading to black in tube; legs yellow, femora and tibiae somewhat darkened with brown, mid and hind tarsi slightly darker; antennae with segments 1 and 2 about concolorous with legs, 1 darker, 2 paler at apex;

3 pale yellowish white in basal sixth, bright lemon yellow in apical two-fifths, darkly shaded with blackish brown between;

4 bright yellow in apical three-fifths, remainder darkened with brown; 5 brown basally and apically, brownish yellow at middle; 6 brown, with a yellowish tinge, particularly at middle; 7 and 8 dark blackish brown. Head 1.2 times as long as wide; cheeks slightly rounded, with three prominent bristles; segment 8 of antennae inserted on ventral surface of apex of 7. Prothorax about 0.6 as long as head; two pairs of bristles at posterior angles moderately long and pointed, others short. Fore tarsus with a long, stout, curved tooth. Wings short, broad, nearly colorless, fore pair with 12 or 13 accessory hairs; subbasal bristles subequal, short, pointed. Tube about equal in length to head, fully 2.5 times as long as basal width and 2.3 times as wide at base as at the abruptly constricted apex.

Barythrips heterocerus sp. nov.

1925]

New N eotropical Thysanoptera

67

St. Thomas, West Indies; C. B. Williams; swept from grass and in epiphytic bromeliad on tree.

Male (brachypterous) . — Very much like female in color and structure, but smaller and slenderer.

I am not perfectly satisfied with the assignment of this species to Barythrips , but have put it here until the female of Barythrips sculpticauda Hood and Williams shall have been made known. Both species agree in most of the important details of structure, and furthermore have the intermediate antennal segments dark at base and pale apically — ah inversion of color pattern which is unusual. The mode of insertion of the eighth antennal segment in heterocerus is of interest.

Pygothrips conifer sp. nov.

Female (macropterous) . — -Length about 1.3 mm. Brown, shading to almost black in segments 8 and 9 of abdomen; tarsi and articulations of legs paler; “tube” bright brownish orange, tipped with black; segments 1 and 2 of antennse clear yellow, 3 yellow at extreme base, remainder of antennae shading to dark blackish brown in last segment; wings light brown, darker at base, with a pale median streak. Head fully as wide as long; eyes about 0.4 as long as head and two-thirds as wide as their interval. Prothorax 0.6 as long as head, all bristles pointed. Wings without accessory hairs on posterior margin. Fore tarsus with a short, strong tooth. “Tube” sub-conical, fully 1.4 times as long as basal width, sides nearly straight.

Trinidad; C. B. Williams; from dead branch of Lagers- troemia infested with bromeliads.

More closely allied to metulicauda Karny from Java, than to rugicauda Hood, the genotype. This and the following are the first species of their group to be recorded from the New World.

Pygothrips nigricauda sp. nov.

Female (macropterous). — Length about 1.7 mm. Cofor dark brown, shading to opaque coal-black in “tube” and distal

68

Psyche

[February

half of abdomen; femora brown at base, shading to bright yellow apically; fore tibine yellow, shaded with brown at sides, mid and hind tibiae darker than femora; tarsi brownish yellow; antennae yellow in segments 1 and 2, 1 lightly shaded with darker at base; 3 yellowish brown, paler at base of pedicel and distinctly lighter in color than rest of antenna, which is nearly black. Head 1.3 times as long as wide, broadest iust behind eyes, thence tapering to base, which is about 0.85 as broad; eyeis about one- fourth as long as head, hardly four-fifths as wide as their interval. Prothorax slightly more than half as long as head, bristles pointed. Wings with 5-7 accessory hairs. Fore femora long, swollen, fore tarsi with a long, stout, curved tooth. “Tube” subconical, about 1 . 36 times as long as basal width, sides slightly arched.

Male (brachypterous). — Slenderer than female, with fore femora greatly swollen and longer than head; tarsal tooth long and curved.

Trinidad; C. B. Williams; from branches.

The long head is distinctive.

Bradythrips Hood and Williams, gen. nov.

((3pa&vs, slow; Opuf/, a wood worm).

Antennae seven-segmented. Vertex of head with one pair of prominent bristles; antennal segments 3-5 decidedly longer than wide; tube much longer than head, very slender, fully ten times as long as greatest width and with four long hairs at tip.

Genotype : Bradythrips hesperus Hood and Williams, sp. nov.

Allied to Urothrips Bagnall by the 7-segmented antennae but abundantly distinguished b}r the other characters given in the diagnosis.

Bradythrips hesperus Hood and Williams sp. nov.

Female (apterous). — Length about 1.8 mm. Color straw yellow, with head (except sides), pterothorax, and middle legs, brown; abdomen at sides, tip of tube, hind legs, and last an- tennal segment, shaded with brown or gray; bright red sub-

1925]

New Neotropical Thysanoptera

69

hypodermal pigmentation in head, ptero thorax, and along sides of abdomen.

British Guiana; C. B. Williams; at base of Imperata caudata.

This and the following are the first species of their super- family to be recorded from the New World.

Stephanothrips occidentalis Hood and Williams, sp. nov.

Female (apterous) .—Length about 1.4 mm. Color straw yellow, with head, prothorax, and fore femora (except apex) dark brown; tube brightened with yellow, shading to dark brown at extreme apex. Vertex of head with three pairs of long bristles, which are slightly expanded at tip, the middle pair about three- fifths as long as third antennal segment, others shorter, outer pair set close to, but at a lower level than, the other. Ninth abdominal segment about 2.6 times as long as eighth. Tube nearly 1 . 5 times as long as head and about 1 . 8 times as long as ninth abdominal segment.

St. Croix and Trinidad; C. B. Williams; from bushes.

Readily known from its congener by the characters given above.

70

Psyche

[February

SOME HITHERTO UNDESCRIBED HABITS OF MESKEA DYSPTERARIA GROTE.

By Richard L. Schwartz,

University of Texas, Austin, Texas.

A number of lepidopterous galls were collected during the summer of 1923 in the stems of Abutilon incanum while the plants were still in foliage. When opened these galls were found to contain lepidopterous larvae of the internal borer type, naked and spotted with brown. The gall chambers were entirely free of excrement, yet there was no apparent opening of the gall. Upon careful inspection, it was discovered that there were several minute holes scattered over the gall and through these holes the tiny pellets of excrement were thrust out by the moth larvae after discharge. Since these early galls were immature, none of them were kept.

On February 8, 1924, about thirty galls were collected from the dry stalks of another species of mallow, Malvaviscus drum- mondii near Austin. From that date on during the winter and early spring more galls were obtained until the total number collected and examined reached sixty-three. In February when the first galls were gathered, photographs were taken of the galls and caterpillars. These are shown in the accompanying figures (la; lb) including one case in which three larvae were so close together that the result was a single gall. All of the galls gathered were on the stems except a single one which was at- tached to a leaf petiole.

Galls were broken open from time to time to observe the time of pupation and the first pupae were found on March 24. In the galls collected in February, the caterpillars had eaten out all the tender pith of the stems, but the hard lignified outer wall of the stem was not penetrated. In all the galls which contained pupae it was found that the larvae had cut through the hard wall a small, round perforation large enough to allow the mature moth to emerge, but the paper-like outer membrane of the stem was intact. The sawdust accumulated by the boring of this hole was piled up in both ends of the gall, and in the upper end was also the last larval skin. The sealed aperture is undoubtedly a

1925] Hitherto U ndescrihed Habits of Meskea dy spier aria

71

Fig. 1. Meskea dyspteraria Grote. FA, galls; B , communal gall formed by three larvae; C,'moth resting on gall; D , mothslin copula.

precaution which permits the easy emergence of the moth, and at the same time prevents the location of the place of emergence by parasites and enemies of the pupae. The moths usually pupate upside down with the head placed directly in front of the future doorway through which the adult is to come out. The position of the pupae in the gall is shown in figure 2b.

On the morning of March 15, 1924, the first moth emerged.

72

Psyche

[February

This was the female which was sketched as shown in figure 2a. No other moths came out till April 20 when two males appeared. These were killed and with the female were sent to Dr. H. G. Dyar, of the United States National Museum at Washington where Dr. Busck kindly identified them as Meskea dyspteraria Grote.

On April 23 three more moths emerged. One of these, was a male. The gall from which it emerged is shown in figure la. Photographic work was done under great difficulties as the

Fig. 2. Meskea dyspteraria Grote. a , Female moth resting on gall ;b, section of gall showing position of pupa; c, male with copulatory tuft extended; d, egg seen from the top; e , same from side.

strong light necessary for the camera disturbed the moths. This difficulty was later overcome in studying the moths at night by the use of a red light which did not seem to disturb them. After April 23 several moths emerged each warm night, although six was the greatest number of living moths under observation at any one time. On June 6, when the author left Austin, several of the moths had not yet appeared.

Some very interesting and hitherto unobserved details of the habits of the moths in mating were observed with the use of the red light mentioned above. Soon after darkness, the male moths extrude from the posterior tip of the abdomen an elongated

1925] Hitherto Undescribed Habits of Meskea dyspteraria 73

tuft of hairs which we may call the copulatory tuft. After this tuft is extruded, it is spread out much in the manner in which a peacock or turkey spreads its tail. This spreading is dorso- ventral as well as lateral, and the tuft is vibrated if the moth is disturbed. Though the presence of the female in the same breeding cage is not necessary to induce this behavior, there every indication that the tuft functions in mating. Unfortunately the author was not able to observe the moths at the moment in which the final stages of courtship took place and a more thorough study would probably bring out other details of interest. In this species the wings and body of the two sexes are identical in color pattern.

Attempts to photograph the copulatory tuft when spread were rather unsuccessful, as the moth usually vibrated the tuft when disturbed or retracted it entirely, and the feathery nature of the tuft makes it a hard object to photograph even when ex- panded. Its form is however shown in a sketch reproduced as Fig. 2c. Fig. Id, is an excellent photograph of the contracted position of the tuft in copulation. The tuft is drawn into the abdomen and remains completely concealed in the daytime.

Several females deposited eggs after copulation; but no larvae were hatched. The odd shaped egg is shown in top and side view in figures 2, d and 2, e.

Young galls found on Abutilon May 24, 1924 showed the first larval instar. The life history has thus been traced except the larval molts.

Meskea dyspteraria was described in 1877 by Grote in the Canadian Entomologist. A very brief and utilitarian life history is given by Heinrich in the Journal of Agricultural Research and his plates give only drawings of the head of the larvae, the pupae, and the arrangement of the larval setae.

Some of the observations made in this paper are not new; but those describing the copulatory tuft in the male are, as far as I have been able to find, original and unique. I intend to make some further study of this species before discussing it in relation to other know phases of insect courtship.

In the photographic; work on this paper I owe much to Reginald Painter, formerly tutor in Zoology at the University

74

Psyche

[February

of Texas and now assistant in entomology at the University of Ohio.

References.

Grote, A. R. A new Species of Geometrse, Canadian Entomolo- gist, Vol. 8, 1877, page 114.

Heinrich, Carl. Some Lepidoptera likely to be Confused with the Pink Bollworm, Journal of Agricultural Research, Vol. 20, 1921, page 828 and plates 101, 104 and 109.

Whatever Your Question

Be it the pronunciation of Bolsheviki or soviet, the spelling of a puzzling word — the meaning of Mighty, fourth arm, etc., this Supreme Authority —

WEBSTER’S NEW INTERNATIONAL DICTIONARY

contains an accurate, final answer. 400,000 Words, 2700 Pages. 6000 Illustrations. Regular and India Paper Editions.

GK Sc O. jVIKl^RI^JVX CO., SPRINGFIELD, MASS.

Write for specimen pages, prices, etc., and FREE Pocket Maps if you name “Psyche",

500 Pin Labels all alike, 50 Gents.

1000 Pin Labels all alike, 80 Cents.

Smallest Type. Pure White Ledger Paper. Not over 4 Lines nor 30 Characters (13 to a line) Additional Characters, 2 cents each, in total and per line, per

500. Trimmed.

Prices subject to change without notice.

C. Y. BLACKBURN, 7 Emerson St., STONEHAM 80, MASS.

CAMBRIDGE ENTOMOLOGICAL CLUB

A regular meeting of the Club is held on the second Tuesday of each month (July, August and September excepted) at 7.45 p. m. at the Bussey Institution, Forest Hills, Boston. The Bussey Institution is one block from the Forest Hills station of both the elevated street cars and the N. Y., N. H. & H. R. R. Entomologists visiting Boston are cordially invited to attend.

We are the Headquarters for Entomological Supplies and Specimens.

Special attention is called to our hand made Smitt boxes.

American Ent. Co, Steel Pins, Klaeger Pins, Life Histories of Beneficial and Injurious Insects, Type, Mimicry and Protective Coloration Collections; also Col- lections of Household, Garden, Orchard,

Forest and Shade Tree Pests. Living Pupae from November to March.

Exotic Insects of different orders; also Biological material for dissection.

Catalogue No. 33, Entomological Supplies, free on application.

WARD’S NATURAL SCIENCE ESTABLISHMENT

84-102 College Ave., Rochester, N. Y.

PSYCHE

A JOURNAL OF ENTOMOLOGY

Established in 1874

VOL. XXXII APRIL, 1925 No. 2

TABLE OF CONTENTS.

Growth of Ant Mounds. E. A. Andrevjs 75

Notes on the Giant Water Bugs. H. B. Hungerford 88

Insect Food Habit Ratios on Quelpart Island. H. B. Weiss 92

Notes and Descriptions of the Cercopidae of Cuba.

Z. P. Metcalf and S'. C. Bruner 95

Pectinate Antennae in the Geometridae. V/m. T. M. Forbes. 106

Note on Panurgidae (Bees). Charles Robertson 113

Another Instance of the Northward Migration of Odonata in the Spring.

R. H. Howe, Jr H3

The North American Varieties of Volucella bombylans Linn.

Charles W. Johnson 114

International Entomological Congress 117

The New York State List of Insects 118

CAMBRIDGE ENTOMOLOGICAL CLUB

OFFICERS FOR 1925

President .	. J. H. Emerton
Vice-President	R. H. Howe
Secretary	. . . . J. Bequaert
Treasurer	. . . . F. H. Walker*
Executive Committee	. A. P. Morse, S. M. Dohanian S. W. Denton
EDITORIAL BOARD OF PSYCHE EDITOR-IN-CHIEF
C. T.	Brues, Harvard University ASSOCIATE EDITORS
C. W. Johnson,	Nathan Banks,
Boston Society of	Natural History. Harvard University.
A. L. Melander,	A. P. Morse,
Washington State	College, Peabody Museum.
J. H. Emerton,	J. G. Needham,
Boston, Mass.	Cornell University. W. M. Wheeler, Harvard University.

PSYCHE is published bi-monthly, the issues appearing in February, April, June, August, October and December. Subscription price, per year, payable in advance: $2.00 to sub- scribers in the United States, Canada or Mexico; foreign postage, 15 cents extra. Single copies, 40 cents.

Cheques and remittances should be addressed to Treasurer, Cambridge Entomological Club, Bussey Institution, Forest Hills, Boston 30. Mass.

Orders for back volumes, missing numbers, notices of change of address, etc., should be sent to Cambridge Entomological Club, Bussey Institution, Forest Hills, Boston, 30, Mass.

IMPORTANT NOTICE TO CONTRIBUTORS.

Manuscripts intended for publication, books intended for review, and other editorial matter, should be addressed to Professor C. T. Brues, Bussey Institution, Forest Hills, Boston, 30 Mass.

Authors contributing articles over 8 printed pages in length will be required to bear a part of the extra expense, for additional pages. This expense will be that of typesetting only, which is about $2.00 per page. The actual cost of preparing cuts for all illustrations must be borne by contributors; the expense for full page plates from line drawings is approximately $5.00 each, and for full page half-tones, $7.50 each: smaller sizes in pro- portion.

AUTHOR’S SEPARATES.

Reprints of articles may be secured by authors, if they are ordered before, or at the time proofs are received for corrections. The cost of these will be furnished by the Editor on application.

Entered as second-class mail matter at the Post Office at Boston, Mass. Acceptance or mailing at special rate of postage provided in Section 1103, Act of October 3, 1917, authorized on June 29, 1918.

PSYCHE

VOL. XXXII. APRIL 1925

No. 2

GROWTH OF ANT MOUNDS.

By E. A. Andrews.

Johns Hopkins University.

It is well known that many sorts of ants dig into the ground and carrying out mouthfuls of earth soon make conspicuous ant hills above their subterranean dwellings. In some, these mounds are of considerable size and of long duration and serve as nests or places for rearing the young.

The best known mound builder in America is the rather large red and black Formica exsectoides, the mound builder of the Alle- ghanies, whose mounds are seen here and there in Nova Scotia, Ontario, Maine, New Hampshire, Massachusetts, Connecticut, New York, New Jersey, Pennsylvania, Maryland, North Caro- lina, Georgia, Wisconsin, Illinois, and Colorado. As one of these mounds may contain more than a ton of earth it becomes of interest to find out how long the ants must labor to accumulate so relatively vast masses of material. One of these ants, alive, weighed 10 milligrams.

When the Rev. H. McCook in 1876 studied the “ant city” of local fame near Holidaysburg in Pennsylvania where this ant, which he called the wood or fallow ant (locally called pismires) had built up as many as 1700 mounds, he soon decided, from the occurrence of the very largest mounds on old charcoal hearths, that even the mounds that might contain 300 cu. ft. of earth were not of any very great age. He also recorded that a field plowed in September 1875 showed new mounds in February 1877; one ten inches high and 35 in diameter; a second 14 by 48. As the work of the ants stops in November these mounds were made, he thinks, in a little over one season of work. Again in a corn field in July 1876 were two new mounds, each made in two months, or one third of a working year, and each measured 8 inches in height and 18 in diameter. The first hill was, he

76

Psyche

[April

points out, built up at the rate of 1/i cu. ft. per annum, the second at the rate of 3, and the other two at the rate of 1. The second hill arrived in about one year to above the average size of the mounds of that region, and its very great rapidity of growth may have been due, McCook supposes, to the fact that it was not a new colony but was reconstruction work built up on the ruins of an old mound.

McCook inferred that mounds might require five to seven years to be built up to average size and that once full grown they do but hold their own. He says there were good reasons to sup- pose that some of the large mounds might be thirty years old but no evidence that any of them last through great periods of time. McCook states that the two small mounds in the corn field were probably new communities and it is evident that only such new communities should be considered in reckoning the initial rate of construction. Each mound is the communal work of a family and grows as the family prospers and multiplies year by year since its individuals escape death in the winter by withdrawing deep under ground in subterranean tunnels and are known to be able to live in captivity as much as seven years. The very rapidly rebuilt mounds on the other hand are the reconstruc- tion work of prosperous communities that are able to repair even great amounts of destruction and removal of old mound materials.

As these estimates of McCook on the rate of building mounds by Formica exsectoides seem to be the only ones printed, the following data collected near Baltimore, Md., may be worthy of record.

Near Lutherville and Timonium in Baltimore County there is a settlement of these ants embracing some two hundred mounds and measurements made at irregular intervals from 1905-1924 furnish added means for judging of the rate of growth and the age of the mounds of these ants. During this period some of the mounds have remained in existence; either growing or remaining stationary in size; others have disintegrated; and others have sprung up anew. Parts of the entire area have been largely abandoned and other parts have been invaded by new mounds. In one of these newly populated areas measurements have been

1925]

Growth of Ant Mounds

77

made of a certain new mound and incipient community which has grown to maturity along with many others established during this period in the same area, which was a very sterile flat of several acres covered with reddish-iron-ore earth: refuse removed formerly from large adjacent bog-iron-ore pits.

When first seen July 7, 1906, this new nest was conspicuous amidst the sparse grass as being a few handfuls of light yellow and white pellets of irregular size brought up from the subsoil and piled up amidst the grass in an irregular mass three inches high and nine inches wide. There were no holes into the middle of the nest, but large irregular entrances about the base on the N. W. S. and E. sides. It was about 150 feet from an old nest, No. 59 of a survey in 1905, that has remained there since despite the encroaching Japanese honeysuckle and has grown from five, to six by eight feet in diameter, and from 2 to nearly three feet in height, with a circumference of 22 feet. Possibly from this large nest came the female that started the incipient nest. When disturbed the ants in this incipient nest swarmed out and made a rustling sound running over the dead leaves near the nest, but they were but few in number though so rapidly covering all the surroundings.

When next measured, November 13, 1906, this incipient nest had grown to be a considerable mass fifteen inches across and four to five in height. This, then, was the maximum of the first year’s building.

After another year, October 24, 1907, this incipient nest measured 6 by 18 inches. The ground on which it was placed was an artificial ridge falling off to the East as a shallow de- pression and from now on it became evident that the earth placed by the ants tended to spread unequally and the measurements from the ground level were greater on the east and less on the west. Thus June 12, 1908, the mass of earth was 21 inches from NW-SE, 18 inches wide from NE-SW, while the height was 8 inches from the ground on the west and 12 from the level on the east. More ants were now busy over the mound, but grass blades and small shoots of honeysuckle were growing up in the midst of the nest. In the fall of that year, October 18, the mound measured 9 by 26 inches and was quite conspicuous an 1

78

Psyche

[April

well covered with small sticks and minute pebbles as well as several snail shells, flat and several millimeters in diameter; thus showing that much material had been added to the mound from surface collections, while originally all the material had come from subterranean excavations. At that date, though there had been frosts, the temperature was 70° F. and yet only one or two ants were to be seen on the mound. No more measure- ments were made till August 7, 1914. It had then become a noticeable mound arising abruptly on the west where bordering an obscure path apparently used by dogs, and sloping gradually on the east down to the lower level. The top of the mound was bald, covered with small sticks and stones, but about the middle height was a tonsure of straggling grass stems. The “stones” were such as ants collect; some may be 7 x 5 mm. and weigh 200 milligrams. Many ants were active about the base, but none up on the mound. A similar nest stood some 80 feet to the North. The measurements taken were: height 12 on west; 19.5 on east; diameters 42 east and west, 38 . 5 north and south.

September 19 of the following year, 1915, it measured 16 on west, 24 on the east, 47 north and south, and 52 east and west. It seemed well cared for with large fragments of stick and stone recently added to the dome to build up after heavy rains and was covered with very active ants. Both base and summit were bare with the sparse grass growing out from the sides of the mound in a zone.

After more than a week of rain in a cold wet season, the nest, June 18, 1916, was still in good form with much fine sand newly applied and swarming with actively working ants. The measure- ments were: west 16, east 21 in height; north and south 50, east and west 53 in width. Thus the mound had not gained, but apparently lost somewhat of its greatest height since the previous autumn, though its increased diameters suggested winter de- nudation had spread the material and as yet the loss of height had not been made good.

No measurements followed till the spring of 1919, April 13, when on account of cold few ants were working, though in the immediate neighborhood some dozen fine new nests of this same general age all showed much recent activity by presence of

1925]

Growth of Ant Mounds

79

fresh earth over the domes. Then as in 1916, the incipient nest remained very much steeper on the north slope and still had an irregular band of green grass in tufts from half way up the slopes to near the base. The surface was covered about an inch deep with fine pellets and dried earth apparently brought out this spring to replace denudation and a series of partly open tunnels and pits across the apex of the mound showed that the interior had been exposed by destruction and removal of the roof in that part. The measurements were: west 15, east 24, north-south 58, east-west 54. Later, October 18, the grass around the middle zone had grown tall but the honeysuckle that surrounded the nest stopped abruptly, leaving a bare narrow yard all about the base of the mound, no spray or stem of the honeysuckle reaching over this dead space to the base of the mound. The bald top with bits of stick and gravel extended down farther on the southerly exposure. The measurements were: west 18, east 28, east-west 60, north-south 63. Thus in this single season of building the ant’s mound had risen three to four inches and spread out five to six in diameter.

July 25, 1920 the mound swarmed with ants but some other mounds near by were more populous. The bald top of the mound was in sharp contrast to the sides grown up with tall grass amidst which the ants had accumulated heaps of earth pellets that formed a spongy mass with irregular cavities. The grass holds the sides in very steep slopes. The moat-like encircling dead space on the level without the base of the mound was strewn with dead leaves but the honeysuckle stopped short at its outer edge. The mound had grown so that the North-west-South-east diameter much exceeded the shortest diameter. The measure- ments were then: West 18-20, East 24-30, North-South 57, East-west 54. By the 19th of September the mound had in- creased somewhat; the height being West 21, East 28, the width North-south 60, East-west 57. Very few ants were to be seen at this time of year and the nest had been got into fine condition for the winter, being covered with fresh roofing of whitish clay pellets as if from deep subsoil, along with many bits of grass stems, one to one and a half inches long, dispersed through with scattered black, dried, excreta of some large caterpillar. At the

80

Psyche

[April

end of its fifteenth year the mound was rather imposing, being 16 feet about the base, with abrupt slope on the north 30 inches up the steep to the summit, and 48 from the summit down to the ground to the South, while the west slope was 33 and the east 44 inches from summit to ground level.

August 21, 1921, the mound was in fine condition with newly capped dome, but owing to cool weather and time of day, 6.30 p. m., but few ants were slowly going over the dome and about the neighborhood of the nest. The measurements were: West 21, East 30, North-south 72, and East- west 72. The greatest North-west-South-east diameter was 76 inches. The surface distances over the top were: East-west 82, North-south 85 and greatest North-west, South-east 88. The circumference of base was 18 feet. In general appearance the mound presented a high state of differentiation. The surrounding bare space or moat strewn with dead leaves and twigs of vine that stops in growth at its outer edge, makes more striking the sudden rise of the mound from the level of the ground. The lower parts of the slopes of the mound are covered with talus of loose mouthfuls of earth rolled down from upper levels. Higher up is a faint zone of scattered grass, and still higher up a higher zone of still more sparse grass. The rounded dome is free from grass and covered with fresh light-colored subsoil and bits of dead twigs deposited to form a sort of rudely thatched roof. Near the summit of the north a minute crater seemed possibly still open to the interior. The north face remains much the steepest and the summit is nearer the north and far from the southerly limits of the mound. At this period of great development of the mound a new in- cipient nest was being constructed 42 inches to the North-west from the adult mound and if this may be regarded as an offspring from the large community, it may indicate its maturity and a successful effort at colonization of the neighborhood.

July 20, 1922, with temperature of 93, after many rains, the mound was swarming with very active ants that also spread many feet from the nest amidst the honeysuckle and grass. The mound presented several open holes of exit near the top but no ants were working near the top. The growth of grass on the sides of the mound was sparse and sickly, the honeysuckle dead

Growth of Ant Mounds

81

192 5

in a broad band of a foot width about the base. The elongation of the mound to the South-east was very evident as well as the great steepness of the north slope. The measurements were: West 23, East 30, North-south 81, East-west 79, as computed from surface measurements of East-west 91/^ and North-west, South-east 99. Measured again in the fall, September 10, 1922, it was: West 20/^, East 32, width North-south 87, East-west 73. The distance around the base of this seventeen year old mound was 20 ft. 6 in., the distance up the north slope 3 feet and up the long south slope 5 ft. 22 in. The tape over the surface showed a distance of 8 feet 4 in. in the North-south direction and 7 ft. 4 in. East-west. This being a clear warm day ants were abundant upon the mound carrying up and dragging earth pellets toward the summit which was conical and closed, in spite of heavy recent showers. The mound was recently covered with fresh light earth with very many light fluffy cast-off pupa cases lying about as if brought up by ants from within the nest. In some parts of the circumference of the base there was more than a foot in width of dead honeysuckle. Large streams of ants ran to and from a tulip tree about twenty feet distant and many up and down the tree. Fifty feet to the east were evident several new young nests.

In the next year, 1923, measurements were made July 8, when the heighth had fallen, west 19, east 29, east-west 76, north-south 84 corresponding to surface distances of 89 and 98 inches. Apparently the lessened height might have come from denudation which added to the diameters. The mound was in fine state of preservation, not injured at the top and the ants were bringing out earth through holes near the top. The sparse grass still present about the upper reaches of the mound did not prevent much fine clean earth from rolling down and spreading to the south-east.

The next measurements, January 27, 1924, showed a greater depression of the summit, west 16, east 27, north-west-south- east 86, north-east-south-west 75. For the first time the angles of slope were measured as follows: North 45°, West 45°, East 40 -38°, longest South-east slope 35°. Other measurements were: circumference 20 ft. 2 in., distance over top North-west-

82

Psyche

[April

south-east 98, northeast-south-west 89. An evident cause for the diminished height was the crushed in state of the top of the dome which was flattened and marked as if by human footprints partly filled in by the ant’s work. No actual break into the in- terior remained. A large human footprint on the north-east slope near the top indicated disturbance with the normal surface. The entire mound was frozen stiff, smooth on the surface with some of the small tufts of grass still green about the lower parts of the slopes.

July 20, 1924, after several days clear and dry, in a very rainy season, the ants were very active over the surface of the mound at four to five p. m. Mound in a fine state of repair with four holes near the top on the north and one at the summit nearly, from which ants emerged ; other holes concealed by talus except on the base of the north where talus was absent and old holes of egress show plainly, so that the north side seemed inac- tive and dead in comparison with the very long talus of fresh mouthfuls running far down to the south-east. Grass about eight inches high but not very flourishing made a ring about the bare summit. The moat or bare space about the base of the mound was very conspicuous from dead defoliated honeysuckle and stone or gravel made clean by the ants having removed the earth. Two dead branches projected from south-east part of east side near the base. Measurements as made with the aid of a level and angle, tape and yard-stick were: height, west 19/4 inches, east 33 inches, diameter east-west 88 inches, north-south 85 inches, north-west-south-east 96 inches, width of moat: north 12, west 19, south 16, east 14, circumference at base: 22 ft. 6 in., circumference of moat 29 ft. 8 in. Angle of slope: north 35°, but at base 45° where dead and full of holes; south- east 28°, west 35°, east 35°. Distance by tape up west side: 3 ft. 4 in., up north 3 ft. 4 in., up east 4 ft. 4 in., up south-east 5 ft. 11 in.

The final measurements made at the end of the year 1924, December 20, showed the mound as it then appeared at the height of its recovered maximum, after the previous period of depression but they are not added to the following table since

1925]

Growth of Ant Mounds

83

MEASUREMENTS

CALCULATIONS

Dates	Height	£	W	Width	CQ fc	E. W.	Circumference in inches	Contents cu. ft.	Increment cu. ft.	Duration months.	Increment j per month.	| Increment per j working month.	| Increment per 1 working day in cu. in.
1906-7-7	3			9				.012	.012	7	.001	.006	.345
1906-11-13	4-5			15				.151	.139	4	.034	.046	2.699
1907-10-24	6			18				.291	.013	11	.001	.028	1.636
1908-6-12		8	12		18	21		.564	.273	7	.039	.136	7.733
1908-10-18	9			26				.912	.348	4	.087	.087	5.011
1909-12-31									.518	14	.037	.103	5.922
1910-12-31									.444	12	.037	.088	5.068
1911-12-31									.444	12	.037	.088	5.068
1912-12-31									.444	12	.037	.088	5.068
1913-12-31									.444	12	.037	.088	5 . 068
1914-8-7		12	19		38	42		3.72	.296	8	.037	.095	5.472
1915-9-19		16	24		47	52		7.35	3.63	13	.279	.726	41.817
1916-6-18		16	21		50	53		7 . 359	.009	9	.001	.004	.230
1917—12-31									.874	18	.048	.006	.345
1918-12-31									.586	12	.048	.009	.518
1919-4-13		15	24		58	54		9.012	.193	4	.048	.193	11.116
1919-10-18		18	28		63	60		13.048	4.036	6	.672	.800	46.08
1920-7-25		/ 18			57	54		10.029	-3 019	9	-.35	-.115	-6.624
		\20	30/
1920-9-19		21	28		60	57	192	12.009	1.980	2	.990	.990	57.024
1921-8-21		21	30		76	72	216	20.945	8.936	11	.812	1.789	102.846
1922-7-20		23	30		81	79		25.44	4.495	11	.408	1.124	64 . 742
1922-9-10		204	32		87	73	246	25.44	0.000	2	.000	0.000	00 . 000
1923-7-8		19	29		84	76		23.04	-2.40	10	-.240	-.800	-46.08
1924-1-27		16	27		86	75	242	20.988	-2 . 042	6	-.340	-1.021	-58.809
1924-7-20		19 H	33		85	88	270	29.813	8.825	6	1.470	2.941	169.401
					(96)

they so closely agree with those of July, some slight falling off being due to a defect near the north summit caused by human feet. This table shows all the measurements taken in the years 1906-1924 inclusive, as well as some calculations of bulk reckoned on the assumption that the mound may be a conical figure. Graphs have been made to illustrate the main facts of the table. They show that the growth in height and in diameter taken from the measurements of the last part of each year, rises steadily for a few years as two nearly parallel curves and then the heighth curve becomes more flat and with indented summit while the width curve continues to a much greater heighth and with no

84

Psyche

[Apri ]

flattening, yet with depressions and recoveries. In a drawing showing the projections of the circumference of the mound as measured or calculated each year the circles enlarge concen- trically and rather uniformly at first and then, in this special case of the mound being erected upon unlevel base, the north- west-south-east diameter elongates and the outlines become more elliptical with the axis shifted to west of north.

The curve of the bulk, as plotted from the calculated bulks at ends of years, shows steady rise for the first years; a de- pression in 1920 followed by greatly increased acceleration with a second set back in 1923 followed by a sharp rise above all previous heights. All these plotted results show irregularities which suggest complex factors acting upon the growth of the mound. The very slow growth the first years is striking as well as the very rapid growth in some later years correlated with the small numbers of ants in the incipient colony at first and the immense numbers in the old successful community. The ants work but half of the year at most, lying dormant in November to March inclusive.

It is to be emphasized that the growth is very irregular, the only constant numbers in the table are the interpolated figures in the seven years when measurements were not actually made. While some of these irregularities are due to crude measure- ments many seem to be the results of complex factors and they actually may express the resultant of causes of success or failure in the community. Thus in the years of no progress and of actual diminution, the loss of height is sometimes compensated by increase in width since the materials are spread out laterally in place of b3ing accumulated at the apex; but there is some- times an actual dimunition in bulk. As the interior of the mound is spongy and readily compressed by large animals walk- ing on the mound it sometimes happens that the measured bulk may diminish when the mass is the same or greater. Possibly some storms may actually carry material in some quantities away from the nest so far that it is not soon returned by the ants and thus the mound may be washed away if the ants are not very successful some seasons in combatting this constant denudation. When a mound is deserted it slowly dwindles

1925]

Growth of Ant Mounds

85

through some years but eventually is all washed away. This particular colony of ants had a very difficult soil to deal with and the progress made at first was much behind that reported by McCook in a region of sandy wood soil. Like all loosely piled earth, the nest must suffer shrinking and condensation from settling when this is not actively counteracted by ant work.

During the years that this mound had been forming others were made in the same soil nearby and these had grown to about the same dimensions. Two little mounds were started near the above mound and may well have been colonial offspring from it. As far as measured they have the same very slow rate of starting and are in the same soil. The following table shows these young mounds starting on their long period of trial, at the same cautious pace:

	Nest No. 2	Width	Contents cu. ft.	Material
1921-8-26	4-5	8-10	•054	Subsoil, clay
1922-7-22	3	1 1	•054	Subsoil, clay
1922-9-10	3	1 1 x 14	.070	Earth
1923-7-8	4	14 x 14	•ii 7	Earth and sticks
1924-1-27	3	15 X 15	. 101	Earth and sticks
1924-7-20	4-5	29 x 17	.421	Subsoil
1924-12-20	5-5 Nest No. 3	18 x 18	.267	Coarse particles and sticks
1921-9-9 1922	2	13 X 7	.030	Subsoil, Sandy
1923-7-8	2	11x5	.019	Subsoil, sandy
1924-7-20	6-5	19 X 16	•045	Clay, subsoil, sticks
1924-12-20	6	20 X l8	.328	Very coarse particles and sticks

The rate of growth under the natural conditions prevailing is but slow, yet when injury is done to the mound the repair work and reconstruction results in very rapid new formation. Thus the little mound No. 2 was cut into two with a saw when frozen, January 27, 1924, and the half carried away, yet by the following July the ants had made good the loss and added to the former height, width and bulk as shown in the table. In the same way a full grown or adult mound cut into two when frozen

86

Psyche

[April

at the same date and one-half removed for study, was found in July completely regenerated . and perfect. The ants can thus accomplish much more work in a given time than they would without the stimulus of destructive injury to the mound.

The actual bulk of the mound at the end of nineteen years of work by the ants of this community is about 30 cu. ft., and was thus accumulated at the average rate of about 2 cu. ft. per year. The table shows the actual slow growth of the first years and the rapid growth of some of the later years. The number of ants is unfortunately not known, but they were very few in the first years and very many indeed in the later years. The nascent community accumulated but few cubic inches per day, the mature colony fifty to one hundred or more. As the single ant is but 1-630 part of a cubic inch the labor done is relatively very great and all the work seemed to be done by individuals without aid of fellows. Such facts led McCook to the estimation that con- sidering the bulk and the speed of construction of the ant mound as compared with the bulk and supposed speed of construction of the pyramids of Egypt, the ant may be much more efficient than man, in fact nearly 700 times as powerful a laborer.

In the building of the mound the first two or three years seem to be exclusively years of mining operations, bringing up the earth from the shafts below ground; but after that period the ants begin to construct the mound from two classes of materials. Not only is the excavation process continued and the removed material added to the pile, but there is more and more, bringing in of surface material, both surface soil and bits of organic matter such as sticks, straws, leaves and other light par- ticles. In the early stages of mound growth the cast up mouth- fuls of subsoil merely accumulate in a loose pile, but the weather compacts them and a denser mass results within which the ants begin to excavate their tunnels above the natural surface of the ground. Thus mound No. 2 had, when four years old, but few internal tunnels merely suggesting the complex labyrinth of the mature mound.

Incidentally it may be noted that in this region no trees seem to be killed by the ants, though that has been described in New England; but the ants keep the Japanese ivy from growing

1925]

Growth of Ant Mounds

87

over the mound and even climb up and kill branches of this vi ne and of catbriar that may project over the mound though some feet above its surface.

Summary.

In a mound of Formica excestoides measured at intervals during nineteen years the growth was not constant but fluctuat- ing in rate. In the earlier years increase was very slow, in the later years very fast. Interruptions in growth and diminution in proportions may to some extent be referred to external inter- ferences, but probably in part to lack of steady success of the community. Other small incipient mounds start at the same slow rate. Mature mounds require many years for completion to full size.

In the first two or three years the material of the mound is mined from the subsoil, but later more surface soil and collected fragments of vegetable matter are added to the pile.

The rate of growth in reconstruction after injury to, or removal of part of, the mound is much greater than the usual rate of growth. Comparison with other mounds suggests the rate of growth to be strongly an individual character of each community under its own complex environment.

88

Psyche

[April

NOTES ON THE GIANT WATER BUGS (Lethocerus and Benacus — Belostomatidae Hemiptera)

By H. B. Hungerford,

University of Kansas.

Two papers have appeared recently on Lethocerus ameri- canus Leidy. One by W. E. Hoffman1 and the other by J. R. de la Torre-Bueno.2 These remind me of some notes that were taken several years ago and may be of interest. In view of the rather wide popular interest in these huge “electric light bugs” it is strange that no one has given us a monograph of even the American forms. Benacus griseus Say and species of Lethocerus often attract attention as they fly about street lights and are sent to the entomologist for determination. A good key for the identification of these forms would be most acceptable.

These notes concern the flight of giant water bugs at Law- rence, Kansas, in May 1920, and some notes on the hatching process in one species. We had occasion to use a large number of these bugs in our studies and, therefore, made some effort to collect them. The notes on the collection of insects about the street lights in May 1920 show periods when the Giant Water Bugs were most abundant in flight.

Most of the collecting was done about two street lights. At 8th and Maine Streets 50 Benacus griseus Say were taken during the evening of May 5th. The bugs were brought to the laboratory alive in a cloth bag. They made a wheezing noise when disturbed and emitted a decidedly fishy odor. It will be noted below that Benacus griseus Say was much more abundant than Lethocerus americanus Leidy.

XW. E. Hoffman, Biological Notes on Lethocerus americanus , Psyche

XXXI, pp. 175-183, 1924-

2J. R. de la Torre-Bueno, The Last Moult in Lethocerus americanus Say Entom. News, Vol. XXXV, p. 369-370, 1924.

1925]

Notes on the Giant Water Bugs

89

Collections of Belostomatidae at street lights in Lawrence, Kansas, May 1920:

Benacus griseus	Lethocerus americanus
Date	Male	Female	Male	Female
May 5th	44	57
May 8th	9	7	1
May 9th	22	29
May 21st	9	12	5
May 22nd	29	16	4	1
Total	113	121	10	1

Some of these bugs were paired and placed in aquaria and numerous egg batches secured. The number of eggs in a mass ranged from 8 to 17 and were placed on supports above the water. In one instance a female laid 17 eggs upon the back of the male. These were attached to the right wing cover near its tip and the male, when discovered, was resting high and dry above the water on the screening of the cage. Most of the egg masses were attacked and sucked dry by the bugs themselves. The eggs when protected from their forebearers have the ap- pearance of those photographed by Dr. J. G. Needham.3 They undergo an astonishing increase in size as they develop. One egg, measured the day it was laid (May 10th), was 4.5 mm. long by 2.25 mm. in diameter. Shortly before it hatched (May 22nd) it measured 6.. 57 mm. long and 2.87 mm. in diameter. The newly deposited egg has the same color as I have noted for Lethocerus uhleri and, if left in the water, does not color up very well. If placed in the air, it becomes longitudinally striped with brown as described by Doctor Needham. The surface is irreg- ularly hexagonally reticulate, the gray and brown being laid down as units. Each hexagonal figure is of a single color, reminding one of the mosaic of a tile floor.

The hatching process is very interesting. I was fortunate

3J. G. Needham, The Eggs of Benacus and Their Hatching, Entom. News, Vol. XVII, p. 1 13, 1907.

90

Psyche

[April

enough to be watching an egg through the binocular when the cap at the cephalic end of the egg popped loose and the nymph began its emergence.4 The cap was forced up by a bubble con- fined by a delicate transparent membrane. After the cap was raised by the bubble-like device the head of the bug slowly advanced into the space delimited by the membrane of the bubble which then burst and rumpled up about the opening of the egg shell. This was not the post-natal molt, for when the bug was nearly out of the shell it was still enshrouded by a delicate garment that embraced each limb separately and was shed as the last rite in the hatching process.

Mr. Hoffman, who studied Lethocerus americanus Leidy in Minnesota, failed to find the eggs in two seasons’ search. On July 20, 1921, at Como Park, St. Paul, Minnesota, I found an egg mass on a dried cattail stalk. The stalk was inclined a few degrees from the perpendicular and on the lower side about six inches above the water surface was found the egg cluster. It consisted of 119 eggs arranged in 6 longitudinal rows and measure- ed 2 inches long and Y2 inch wide. The cattail stalk was brought to the laboratory and placed in an aquarium jar. The following morning several eggs had hatched and some were in the process of hatching. At this stage the photograph submitted herewith was made. The nymphs had the groove in the anterior femora which is characteristic of the genus Lethocerus and in all prob- ability they belonged to the species L. americanus Leidy.

4J. H. Fabre, Etudes sur l’instinct et les moeurs des insectes, Souvenirs Entomologiq-ues, 18 serie, p. 99, 1903. Gives charming account of emerging of Reduvius personatus.

PSYCHE 1925

VOL. XXXII, PLATE II

HUNGERFORD: EGGS OF BELOSTOMATIDAE

. : ... • —

1925]

Notes on the Giant Water Bugs

91

Explanation of Plate II.

Eggs of Belostomatidce .

Fig. 1. Dorsal view of male of Belostoma flumineum Say bearing eggs.

Fig. 2. Egg cluster of Lethocerus americanus Leidy a little en- larged. Note the hatching bug not yet out of its post- natal molt and the one that has completely emerged. There were 119 eggs in this cluster, arranged roughly in 6 rows. Each egg measured 4.5 mm. long and 2.25 mm. in diameter. The eggs are gray with the free or cephalic end blotched with brown, more especially on the exposed side of the egg which is on the ventral side of the embryo as it emerges. The micropylar area is elongate, light in color and surrounded by a brown band. Pale streaks radiate from this area arranging the irregular brown spots in more or less definite rows. Compare the markings of these eggs with that of the one on the right.

Fig. 3. Hatching egg of Benacus griseus Say : (a) The micropylar area, (b) Indicates the space on the head between the eyes that pulsates sharply and at irregular intervals. Appears to have something to do with the inflation of the bubble which lifts the egg cap. (c) The eye. (d) The thin transparent membrane that confines the bubble, (e) The space not yet filled by the advancing embryo. Soon the bug occupies this space and disrupts the mem- brane (d) which crumples about the egg shell at (f). Copied by Miss Kathleen Doering from a pencil sketch by the writer.

92

Psyche

[April

INSECT FOOD HABIT RATIOS ON QUELPART

ISLAND1

By Harry B. Weiss.

New Brunswick, N. J.

The following notes represent an attempt to reduce the activities of the insects on Quelpart Island to certain food-habit types and to express the relative importance of these types in terms of parts of the whole, thus establishing a series of ratios. A list of the insects of this island was published recently by Hanjiro Okamotoas Volume 1, No. 2, Bulletin of the Agricultural Experiment Station, Government-General of Chosen, Suigen, Corea, Japan, March, 1924, and the species in this list were arranged and tabulated in accordance with the predominating larval habits of their families.

According to Mr. Okamoto, Quelpart Island is the most southern point in Corea, being the largest island adjacent to that part of the mainland. Its total area is given as about 718 square miles. The highest point is Mt. Kanra with an elevation of 2056 m., and the island consists mainly of tertiary volcanic peaks surrounding Mt. Kanra with a gradual slope to the sea. Mr. Okamoto quotes Dr. Nakai concerning the native plants which number about 1300 species and it is possible to separate the island into seven zones, “of which the southern limit of each is much higher in elevation than the northern, except at the top, where no relation exists with respect to the ocean current, and the seventh zone is consequently level on all sides.” The flora of the four lower zones is temperate while a more northern flora is found in the three upper ones. The island therefore has a flora of a wide range. A more detailed description of the area can be found in Mr. Okamoto’s paper.

Some 527 species are listed and although this figure does not represent a “complete count,” from the information given as to the routes and times of the collecting trips, it does appear as if it might be considered as a representative sample. Even

Tormer papers on the ratios of insect food habits were published in the Ohio Journal of Science, vol. xxiv, pp. 100-106, Entomological News, vol. xxxv, pp. 362-364 and the Proceedings of the Biological Society of Washington, vol. 38, pp, 1-4.

1925] Insect Food Habit Ratios on Quelpart Island

93

though not strictly representative, it is the nearest approach to it that is available at this time.

In accordance with their family food habits, the 527 species can be grouped approximately as follows:

	No. species	Phyto- phagous	Sapro- phagous	Harpac- to- phagous	Parasitic	Pollen feeders, misc. sp.
Quelpart Island	527	60%	16%	20.8%	1-5%	i-7%

A comparison of the above ratios with the ratios for other sections which have been collected over more thoroughly and which have been treated in a similar way, shows that the para- sitic figure for Quelpart Island is too low, due to the fact that the parasitic Hymenoptera are not represented as they should be. From studies of “samples” in other areas, it appears that the parasitic figure should be about 10 and in view of this, 50 species have been provisionally added to the parasitic group making the total number of species 577 and resulting in the adjusted set of ratios found below.

	No. species	Phyto- phagous	Sapro- phagous	Harpac- to- phagous	Parasitic	Pollen feeders, misc. sp.
Quelpart Island	577	55%	15%	19%	10%	1%

It is now possible to compare these ratios with those of other areas and this has been done in the following table I. It will be noted that the figures in each column do not differ widely and suggest a fixed relationship or at least a close resemblance. The same relationships are brought out in a slightly different manner in table II wherein the comparisons are made with the parasitic food habit as a base. For example, if the adjusted parasitic food habit in Quelpart Island is represented by 1, then the relative importance of the other types is, phytophagous 5.5, saprophagous 1.5, etc.

94

Psyche

[April

TABLE I

	No. species	Phyto- phagous	Sapro- phagous	Harpac- to- phagous	Parasitic	Pollen feeders, misc. sp.
Quelpart Island	577	55%	15%	19%	10%	1%
Western Arctic Coast of N. A.	400	47%	27%	14%	10%	2%
State of N. J.	10,500	49%	19%	16%	12%	4%
State of Conn.	00	52%	19%	16%	10%	3%
TABLE II
Quelpart Island	577	5-5%	1-5%	1-9%	1.0%	0.10%
Western Arctic Coast of N. A.	400	4-7%	2.7%	i-4%	1.0%	0.20%
State of N. J.	10,500	4-i%	1.6%	i-3%	1.0%	0-33%
State of Conn.	6,781	5-2%	i-9%	1.6%	1.0%	0.30%

The adjusted figures for Quelpart Island, with its varied flora, tend to support the suggestion advanced in previous papers, that the ratios between the various types of food habits, based on the species present, vary but little when large areas, each em- bodying different types of vegetation are considered in toto and when the numerical ratios between the species present and the factors tending to reduce or change their numbers are con- sidered as constant.

1925] Notes and Descriptions of the Cercopidce of Cuba

95

NOTES AND DESCRIPTIONS OF THE CERCOPIDdE OF

CUBA.

By Z. P. Metcalf, North Carolina State College, Raleigh, N. C.

and

S. C. Bruner, Estacion Experimental Agronomica de Cuba,

Santiago de las Vegas.

This is the first of a series of papers on the Homoptera of Cuba which have been collected by the junior author during the past eight years. For some unknown reason the homopterous fauna of Cuba has been neglected since the publication of de la

Sagra’s “Historia de Cuba,” in 1856, which

contained a list of the Homoptera of Cuba by Guerin-Meneville, with descriptions of new species, but he did not include a single Cercopid. In 1864, Uhler, (Proc. Ent. Soc. Philadelphia II : 155- 162) described some new species of Hemiptera including three new species of Cecropidse, Clastoptera stolida, Clastoptera undulata, and Tomapsis ( Monecphora ) fraterna, from Cuba. Again in 1876, (Bui. U. S. Geol. Survey I: 348) he described, Lepyronia an - gulifera, from Florida and stated that he had specimens from Cuba. The present list records fifteen species and one variety; twelve species appear to be new to science, indicating very clearly how much the Cercopid fauna of Cuba has been neglec- ted. One new genus is also described, indicating that Cuba may be an important center for the development of this group. No less than five new species of the genus Leocomia Ball are recognized, indicating that Cuba may be an important center for the development of this genus, which was described from Hayti.