1886.] Larva of Sniermthus ocellatns and its Food-plants. 145 



were fed upon apple. On August 16th two of the eight large larvae 

 were quite clearly, though slightly, on the yellowish side of inter- 

 mediate. On August 18th four ceased feeding, and on August 23rd 

 three more without change of colour. The other larva and three 

 small ones died. 



16. Salix triandra (without the whitish bloom upon the under side 

 of the leaves). — The bloom was fur the first part of the time rubbed 

 off with the moistened fingers, but afterwards a tree was found near 

 the Oxford University parks, of which all the leaves were without 

 the bloom, and the larvae were fed upon this food for the later part of 

 their lives. The following results afford a very interesting comparison 

 with those given above, following the use of the ordinary leaves of 

 S. triandra which presumably tend less towards the yellowish varietv 

 of larva than those supplied in the present instance. 



Ten larvae (hatched July 13 — 14th) were variously advanced in the 

 last stage on August 18th, and on comparing them with those (see 

 above) fed upon ordinary S. triandra (most of which were rather 

 older), it appeared that the former would be considerably yellower 

 when they had reached an equal age. There were also other younger 

 larvae upon the same leaves, of which the tendency could be better 

 estimated at a later date. On August 23rd one had ceased to feed, 

 and was distinctly on the yellowish side of intermediate. On August 

 27th the small ones had died without any results, while four of the 

 older ones were full fed, and the others dead (although old enough to 

 indicate what their colour would have been). The results were very 

 uniform, all being on the yellowish side of intermediate, while only a 

 small proportion of those fed upon ordinary S. triandra were at all 

 beyond the intermediate form. 



17. Salixbabylouica. — Four larvae (hatch ed July 10 — Llth) were well 

 in the last stage on August 12th when they were examined. They had 

 been fed on S. triandra for twenty-four hours (July 23rd — 24th) 

 because I was travelling and could not get the proper food. On 

 August 12th they were all well on the yellow side of intermediate: 

 they were again examined on August 20th and 27th, when two ceased 

 feeding, one was practically mature, and one had died. The colour 

 remained the same in all cases. 



When examining these larvae at an earlier date (August 3rd), when 

 they were more numerous, I noticed one which was in the fourth 

 stage, and which possessed the upper row of rust-coloured spots which 

 are often found on the yellow varieties of these larvae. The spots were 

 present on the second thoracic segment (very faintly), and upon the 

 first five and the seventh abdominal segment. To my great surprise 

 I observed that the larva was distinctly whitish, and as I was most 

 anxious to prove that such varities can bear the spots I changed its 

 food from S. bahylonica to apple (August 3rd). The next day the 



