178 THE ENTOMOLOGIST’ S RECORD. 
That represented in plate vi., fig. 5, is a very similar specimen, 
but here the base of the larval lev is clearly seen to be uninjured, 
whilst no trace of the leg itself can be confidently recognised. In the 
imago the leg parts are complete, though they are very much smaller 
than on the opposite side, the terminal tarsal joints being very small 
and only one claw being clearly visible. 
I take it that in fig. 4 the amputation of the larval lee was 
made so high up as to injure without destroying the regenerative 
centre, whilst in fig. 5 this centre remained intact, and we see in 
the imago how much it was possible for it to accomplish in the one 
moult from larva to pupa. J! imagine that the moult from pupa to 
imago does very little to the further development, the pupa being in 
reality rather an immature imago than a previous stage in the definite 
sense in which the larva is so, and, therefore, the development of a 
regenerated part may take place at the pupal moult, but not at the 
imaginal. 
No. 9 (an example not figured) 1s of some interest, as in this 
case the amount of larval regeneration that has taken place is much 
less than in pl. iy., fig. 1, yet the imaginal parts are not much less fully 
developed than in that specimen. The larval leg is represented by 
basal parts perfect, femur (first joint) well represented, but further 
joints are represented only by some wrinkled chitin at its extremity. 
This is not, however, mere crust or scab, but is obviously tissue, 
formed at a moult that has occurred since the injury, and no doubt, 
therefore, does really represent the further joints, and may, when the 
larva was alive, have been more distended, and really showed the 
several parts, but being soft has collapsed at moult, and not un- 
ravelled itself in my manipulations. 
In plate vi., fig. 6, amputation preceded the last larval moult, 
the leg base is normal, and the leg itself represented by a small capsule 
that has the base of a larval femur very recognisable as its only very 
definite feature. On Gonin’s view we ought here to have a well- 
developed leg. We have, however, a very small and somewhat 
anomalous structure. The trochanter, as usual, is normal, and the 
tarsus is complete though very small; but the femur and tibia are 
represented by one curved piece, basally, no doubt, femur, apically 
tibia, both because it possesses a tibial spur and because it articulates 
with the tarsus. The specimen of the larval skin has the lee-piece 
folded under, and so it is not obvious at once as in some of the other 
specimens that have been more successfully displayed. 
These specimens demonstrate that on a larval leg bemg amputated 
there arises from the structures at its base a new larval leg. At first 
of very small size, and with the several parts represented by very smal] 
chitinous scraps, but still often perfectly recognisable as the ‘several 
parts of a complete leg. If this process began early enough in larval 
life, no doubt by the last larval instar a fully formed larval leg would 
be reproduced, practically identical with that of the other side. This 
I have still, however, to prove experimentally, but what we find 
proved is, that if by the last larval instar a leg has been reproduced 
with all its parts, no matter how small they may be, then the imago 
possesses a perfect limb, though smaller in size than its neighbour. 
If the larval leg at the last instar is imperfect as to parts, then the 
regenerated leg will be extremely small, the parts being as it were 
