RELATIONSHIP BETWEEN LARVAL AND IMAGINAL LEGS OF LEPIDOPTERA. 179 



outlined but not fully grown, just as the larval leg is at its first 

 appearance when reproduced after amputation. There is another 

 alternative, and that is that the imaginal leg may be wanting beyond 

 the trochanter. My experiments are too few to determine which 

 distinguishes these cases. 



If the parts are entirely wanting in the imago, it is clear that the 

 larval parts that correspond to the imaginal part, whether as partially 

 formed or merely germinal material, must have been removed. And, 

 further, that the mere basal germinal material that gives rise to the 

 regeneration of the limb, has either been removed, or, more likely, has 

 not been afforded its proper stimulus. In the other case, if the 

 imaginal limb is present, but of reduced size, it is clear that the 

 proper larval source of the imaginal limb has been removed, and that 

 it has been re-supplied by regeneration. 



If certain portions are of full size, either the larval parts to which 

 they correspond have not been removed, or they have been removed 

 so early in larval life that regeneration has had time to become 

 complete. 



It may l^e said that if regeneration can re-form the whole limb from 

 germinal material at the base of the limb, then it is obvious that 

 Gonin's hypothesis of germinal material existing there is admitted. 

 But this is not so. Gonin's hypothesis is that the germinal material 

 for the leg exists there, and that it is employed to form the imaginal 

 limb in ordinary. Whereas the true hypothesis is that this regenera- 

 tion plasm is not used in any way, and does not develop into anything, 

 unless required for its proper function of supplying an amputated 

 limb. This becomes evident if we consider the case of a lobster or a 

 cockroach, where regeneration takes place when a limb is lost, but 

 where there can be no doubt the limb is not grown from the base de 

 novo at each moult, in ordinary circumstances. 



I think we may reasonably conclude from the facts observed : 

 (1) That the old idea, that the larval leg u the imaginal leg, is sub- 

 stantially correct, and that the result of the comparison with trichop- 

 terous and other larvte, showing that the parts present in lepidopterous 

 larvce are the femur, tibia, and tarsus, is supported by them. (2) That 

 there is germinal material at the base of tlae larval leg, which would 

 under ordinary circumstances be functionless, but on the loss of the 

 leg by accident or injury, comes into action and reproduces the lost 

 limb, larval or imaginal. as the case may be, but the reproduced limb 

 is a diminutive sketch of the lost limb, and can only reach a size to be 

 functional after further moults. (3) That the progress made in the 

 regeneration of the lost part is disclosed at each moult, remaining as 

 it w^ere suspended in the intervals ; but there is little doubt that its 

 real progress takes place during these apparently quiescent intervals. 



It Avill, I think, be safe to conclude that M. Gonin's hypothesis is 

 quite erroneous. 



Explanation of Plate VI. 



The Plate represents the third pair of legs of certain imagines of 

 Vorthetria diapar, accompanied by the same legs of the larval skin cast 

 at the pupal moult, the left leg having been removed during the last 

 larval or some preceding instar. In fig. 3 the right larval leg has also 

 been removed. Further details in text. 



