104 ])r. T. A. (^^pinan on Calloinlirys avis. 
and nine on the other being not uncommon, a larger 
number, twelve or thirteen, is rare; they alternate large 
and small, with small ones at each end, sometimes the end 
of each series is smaller than the middle ones. 
In the last instar they are much more numerous but 
much alike and variable in the same wa}'’ in the two 
species ; twenty-one to twenty-three is about the average ; 
one may find twenty-three on one pad and seventeen on 
the other, and as many as twenty-five occur—sometimes 
the front, sometimes the posterior, pad has the fewer 
crochets. They are alternately longer and shorter, but the 
longer set often varies again into longer and shorter. 
In the third instar (Plate XXXII) the hairs are again 
just appreciably longer and stronger in C. ruhi (Plate 
XXXIII). The honey-gland (Plate XXXVI, fig. 2) is now 
quite distinct in both species by a flight of lenticles 
along both lips and by the recession of the principal 
hairs. 
In the last (fourth) instar, the honey-gland (Plate 
XXXVII) is very obvious; a space is made for it by the 
loss of the longer dorsal hairs. The appearance shown in 
Plate XXXVII, fig. 1, of four circles right down in the 
fjland, is one I am familiar with in the larvae of “ Blues,” 
but what precisely they represent, functionally or homo¬ 
logically, I do not know. 
I suspect them, however, to represent the two pairs of 
dorsal hairs (1 and II) changed into glandular structures. 
The hairs at all stages (if we except the curving in the 
first stage) are simple and straight and armed with fine 
spicules; in the last stage, however, these have so 
dwindled that it may be said they have disappeared, 
traces of them are more easily detected in C. ruhi than in 
C. avis. 
At all stages the bases of the hairs are spread out and 
divided by lines, so as to take the floral aspect they have 
in many other Theclids. 
I have already noted the extra hairs that C. ruhi has on 
the prothoracic plate in the first instar. In the later 
instars (Plates XXXIV, XXXV, and XXXVI) the hairs and 
lenticles of the 2 :»rothoracic plate vary so much that it is 
difficidt to find two individuals with precisely the same 
dispositions, so that the differences one may note in a few 
specimens between C. avis and C. ruhi are more ^mobably 
individual than specific. The special angular hair does 
