96 ANTHROCERA TRIFOLII. 



prominent. And seeing them in such a dress at this 

 time, one conld not help thinking that it was meant to 

 assimilate them whilst torpid to the withered stems of 

 their old food plants. Some time in February they began 

 to part company " (earlier than the date recorded in 

 1873), " and to feed again, and not long after, having 

 moulted, they appeared in a greenish-grey coat, with 

 a row of pale primrose spots on each side of the dorsal 

 rows of black X-like spots. Finally the ground colour 

 became much more vivid, either of a bluish-green or 

 else of a rich yellow-green ; but the rows of black 

 spots on the back retained their X-like form, and this 

 seems to be one of the chief distinctions by which this 

 species may be known from filipendulce, these dorsal 

 rows in the latter being formed of a broad and a narrow 

 black spot alternately." 



And I also noted that the Anthrocera larvae effect a 

 moult, not by creeping out of the old skin through a 

 split in front, but simply by standing still, and burst- 

 ing it open all down the middle of the back. This 

 year I tried to compare larvae of trifolii and filipendulce 

 very carefully when full grown, but I cannot describe 

 the forms represented by figs. 2 cand 2d; and, indeed, 

 I made filipendulce the subject of my full descrip- 

 tion, only noting for trifolii such points of distinction 

 as I thought I could detect, and some of these points 

 may be seen in figs. 2 a and 2 b. The form and struc- 

 ture are very like those of filipendulce, the size rather 

 less, the ground colour perhaps more yellow, the large 

 upper rows of black spots are more even ; when the 

 segmental divisions are hidden these spots are like x's, 

 as shown in 2 a and 2 b, and the spots in the lower 

 rows are also smaller than in filipendulce. But it is 

 evident that Mr. Buckler had other varieties, and I 

 have no doubt he could have given more peculiarities 

 for each species. 



The cocoon is slightly shorter than that of filipen- 

 dulce, of the same substance, most generally yellow in 

 colour, but I have one cocoon quite white. The pupse 



