45 



in the case of any specimen with history unknown that the malfor- 

 mation is the result of an error in the regenerative process, and that 

 it is not a vagary of the primary embryonic material. I believe I am 

 not wrong in saying that the latter hypothesis is the one almost 

 universally adopted. My own belief is that it is comparatively 

 rarely that these reduplications date from the original embryonic 

 origin of the individual, and that in the great majority of 

 cases they are the results of errors in regeneration. The normal 

 result of regeneration is the reproduction of the lost part, of 

 normal size if time allows, diminished in size if not, the 

 struggle being to produce all the normal structure by the time the 

 imaginal state is reached, after which, of course, nothing can be 

 done. This statement is broadly true, notwithstanding well-proved 

 facts to the contrary, as, for example, that certain Crustaceans cannot 

 re-develop an eye and its stalk, but replace it by an antenna, due 

 probably to the eye containing offsets from the central nervous 

 system, not capable of being replaced from embryonic cells peri- 

 pherally preserved for purposes of regeneration ; or, again, the less 

 easily comprehensible case of Blattidce. that reproduce the tarsus 

 with one joint missing. 



Such exceptions present in their particular cases the normal 

 process of regeneration, and not only cannot be regarded as terato- 

 logical, but leave unaffected the general principle that regeneration 

 results in the appearance of no damage having been sustained or any 

 alteration having occurred. 



I made certain experiments in 1899 on the legs of Lyjiianfria 

 dispai\ some of which are reported in the " Ent. Rec." for 1900. I 

 made a much larger number of more complete experiments on the 

 same species a year or two after, but I have never reported any 

 of these, entirely — I think, I must confess, from sheer laziness. 



These experiments, as bearing on the present portion of our dis- 

 cussion, showed that when the injury was not too great, and occurred 

 early enough in the larval life, it was the rule for the imago to show 

 very little trace of the injury, but in some cases there was some 

 deformity, and in two or three this deformity took the form of a 

 tendency to duplication of a portion of the limb. One other obser- 

 vation is here of importance : in rearing together a considerable 

 number of larvae, in fact a brood of T. promiba, I was rather sur- 

 prised to find that nearly all were without portions of, generally, several 

 of their legs ; in one or two all the legs were affected ; they had 

 obviously, in the crowded state in which I had kept them, bitten 

 each other to this extent. The legs of the imagines, however, 

 showed no very appreciable defects. I have once or twice found 

 wild larvae with damages to their legs. The point I wish to make 

 here is that, as a natural occurrence, larvae no doubt often have 

 partial amputations of their limbs, and that in regeneration a by no 

 means inappreciable percentage develop duplicated limbs — certainly 

 quite enough to account for the number of such malformations as 



