466 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



habit of using its larvae as implements wherewith to spin its nests, 

 and its queer and varied acquaintances, is the most remarkable of 

 all. 



Commenting on Mr. Dodd's account of the habits of Liphyra 

 hra^solis, Mr. E. Ernest Green noted that the fact that the larvae 

 are really carnivorous seems to be proved by their having seized and 

 attempted to eat some of the ant grubs; but they do not appear to 

 have been satisfied with that diet. He therefore asked, is it not 

 possible that their proper food may be some coccid inclosed in the 

 ants' nests? Mr. Green remarked that in Ceylon the arboreal nests 

 of the same ant almost invariably include colonies of Coccidae, 

 Aphidce, or Aleurodidse. 



Dr. T. A. Chapman described in detail three larvae of Liphyra 

 hrassolis sent to England by Mr. Dodd. Two of these were appar- 

 ently full grown, while the other was very much smaller. In these 

 two sizes the larvae are so very different that had they come from 

 different places and possessed different histories one would never 

 suspect that they were at all related. 



The smallest caterpillar was 6 mm. long and 2.3 mm. broad and 

 very flat, reminding one very much of the larva of Camponiscus 

 luridiventiis (a common alder saw-fly) in its general size and form; 

 or, for that matter, of an ordinary LycoR'na larva, if we make it 

 first, colorless, then nearly flat instead of raised along the dorsal line, 

 and, thirdly, if we somewhat exaggerate the rounded lateral projec- 

 tions of the segmental margins. 



The segmental divisions as seen dorsally are 12, of which the 

 first and last are terminal. White and soft as this larva looks, the 

 margins, nevertheless, have something of the appearance and struc- 

 ture of the full-grown larva, having a strong chitinous binding 

 divided into small cells. 



The prothorax has a slight angle on either side, allowing the 

 front between the two angles to be a transverse straight line. The 

 head is beneath this, placed quite centrally, quite free from it, and 

 capable of much movement, including probable protrusion in front. 



The last segment visible dorsallj'^ is the ninth abdominal, and 

 beneath this is the tenth carrying the anal prolegs. In a ventral 

 view of the larva the true segmental divisions are very plain. 



The prothoracic spiracle lies in the fold between the pro- and 

 mesothorax, but the abdominal spiracles are each on about the 

 middle of its own segment, placed dorsally, about half way from 

 the middle line to the margin. Doctor Chapman did not detect 

 anywhere any hairs or tubercles. 



The nervous ganglia are very visible as reddish brown masses. 

 There is a large one in the prothorax that represents the cephalic 

 and oesophageal. The prothoracic proj)er is placed toward the pos- 



