1898.] APPENDAGES IN THE ARTHBOPODA. 933 



when expressed in a similar manner to the above. [The tarsi in 

 this latter case were not from the third pair of legs only, but this 

 ■would not appreciably affect the validity of the comparison.] 

 Among adults there are occasional cases in which the difference is 

 much greater than 3'5 per cent., and such probably indicate that 

 loss and reproduction have occurred in the later instars. It was 

 noticed that the disproportion between the reproduced tarsus and 

 its normal fellow was somewhat greater in adults of Sti/hpi/r/a 

 orientalis than in adults of Periplaneta americana or P. australasur. 

 This may be due to specific differences in the rate of growth of the 

 reproduced structures, or else to some special liability in the first- 

 named species to accidental loss during the later instars. As in 

 Cockroaches the reproduced tarsus has only four joints, it follows 

 that in cases where one tarsus is normal and its fellow reproduced 

 and the two are of approximately equal lengths, the mean lengths 

 of the joints of the latter are on the \^ hole greater than those of 

 the former. In the case of insects with "complete metamorphosis " 

 reproduced appendages in the imago have always been described 

 as smaller than the normal, though it is possible that they may 

 sometimes become symmetrical, for Newport (61) found that in 

 Vanessa larvae reproduction of a lerf commenced two stadia before 

 pupation \A'as accompanied by progressive increase in size. 



The Steuctitke of the Eepbodxtced Legs in the BLATTiDiE. 



In addition to the general observations on the natural history 

 of ecdysis recorded in my previous paper, the following facts were 

 noticed during the experiments made in the course of the enquiry 

 into the regeneration of the legs in Cockroaches. 



(a) The Length of the Period between Mutilation and Reproduction. 



As already recorded, the legs of 833 nymphs of Stylopyga orientalis 

 were mutilated in various parts and the animals kept in confinement 

 in order that the reproduction of the injured limbs might be 

 observed. In 625 cases out of 1473 mutilations, reproduction 

 occurred. The instances tabulated (see p. 934) indicate the shortest 

 periods which elapsed between mutilation and reproduction in 

 different degrees. [The term " reproduction " implies in all cases 

 that regrowth of the mutilated or amputated parts took place with 

 the tarsus in a /ozw-jointed condition.] 



The total number of cases in which it was possible to keep an 

 exact account of the number of days between mutilation and the 

 ecdysis succeeding, and from which the instances recorded in the 

 table were taken, was hardly large enough to permit more than a 

 mere suggestion that reproduction may take place in a shorter time 

 in early nymphs than in nymphs approaching maturity, and that it 

 may take place among the latter more rapidly in males than in 

 females. There is, however, some definite evidence that repro- 

 duction of the tarsus alone may occur within a shorter time than 

 that of the more proximal regions of the leg. It is probable that in 



