926 MS. H. H. BBINDLEY ON EEPEODUCED [DeC. 13, 



loss, but the dititribution of possible starting-points of reproduction 

 is particulate, being in all cases controlled by the effects of injury 

 or amputation on the portion left attached to the body. Thus in 

 cases where autotomy at a particular region is the invariable result 

 of injury or amputation in any more distal region, it is obvious 

 that the starting-point of reproduction is localized to the spot 

 where the autotomous break occurs. On the other hand, there are 

 cases in which reproduction seems to commence from the actual 

 extremity left on amputation, so that there is no clear localization 

 of the reproductive power. 



There are also certain conditions intermediate between the 

 above extremes. As there is such an obvious relationship between 

 autotomy and the reproduction of lost parts, it is unfortunate that 

 our knowledge of the factors controlling the former event should 

 be so imperfect. The term autotomy has received what seems 

 an undesirably wide application by Giard (35) in an attempt to 

 classify many different phenomena under this title, but in what 

 follows its use is confined to the sudden separation of an appendage 

 or part of an appendage in obvious response to an external 

 stimulus. That the event is truly a reflex action, and therefore 

 dependent on the integrity of the nervous system, has been shown 

 experimentally for the legs of several genera of Decapod Crus- 

 taceans by the exhaustive experiments by Fredericq (26, 27, 28), 

 and for the legs of Locusta by Fredericq (29, 30) and Contejean (19). 

 In the case of other Arthropods the phenomenon is probably 

 of essentially similar nature. But when a series of experiments 

 is conducted on a particular appendage of a single species, the 

 general experience has been that the latent period which elapses 

 between the application of the stimulus and the rupture varies 

 within wide limits ; and not only is this the case, but it is also 

 usually found that though a certain degree of injury or amputa- 

 tion in the distal portion of an appendage liable to autotomous 

 loss will almost invariably bring about quite readily the loss of 

 the remaining portion, it does not ahvays do so, though the 

 stump may break away at the usual seat of autotomy after the 

 lapse of a" day or two. So much has been said to indicate that 

 it is after all not at present possible to draw any satisfactory 

 distinction between autotomy in the strict sense and the dropping 

 away of the stump of an appendage some time after the loss of 

 the distal portion, which latter event is of common occurrence in 

 some forms which do not ever exhibit autotomy as defined above. 

 And a satisfactory conception of autotomy is also rendered difficult 

 by the fact that in cases where its occurrence is characteristic, 

 even what are apparently the most favourable conditions will 

 sometimes fail to induce it. The whole subject offers much 

 difficulty, but enough has been said to show that it cannot be left 

 out of account in connection with the i-eproduction of appendages. 

 Ckustacea. — Among Arthropods autotomy and subsequent 

 reproduction have received the greatest degree of attention in the 

 case of the thoracic limbs of Decapod Crustacea. On the authority 

 of Hallez (45) it appears that the earliest observations in detail are 



