116 M. E. Bordaffe on fAe 



t>^ 



appears in the case of tlie Blattidae, but nevertheless by no 

 means so fast as I have found it to take ])lace in the Phasmid^. 

 The tarsus of a regenerated limb is always tetramerous, 

 and the relative size of its joints is as constant as in. the 

 normal pentamerous structure. The same thing holds good 

 also in the case of the Phasmidse and Blattidas. 



II. It is important to note, in the three families of the 

 Orthoptera pentamera, the way in which the replacing limb 

 grows. Instead of developing freely and in a rectilinear 

 manner at the surface formed by the mutilation, the limb 

 must continue to grow, until the next moult, under the skin, 

 which soon covers the wound. The skin is very thin and 

 not at all chitinized, consequently retaining a certain elasticity 

 and transparency even in the Phasmidae. The limb in pro- 

 cess of regeneration barely makes a projection under the skin 

 or is so little apparent that in most cases it needs careful^ 

 examination to determine its presence. 



The young limb in order to develop under these conditions 

 is obHged to coil upon itself and assume a spiral form *. It 

 is then not seen until the next moult sets it at liberty. Wiien 

 it appears it is in the form of a little blackish appendage, 

 hardly a millimetre or two in length, a smallish limb which 

 immediately unrolls, becoming turgescent and rectilinear. 

 From black the limb becomes rapidly of the customary 

 yellowish green, except in the Blattidte and certain Phasmidie, 

 which are of a brown colour. 



These changes take place under observation with a rapidity 

 which is really marvellous, and comparable to that which we 

 see in the develoj)ment or, rather, extension of the wings, 

 especially in the Lepidoptera, when the perfect insect has just 

 left the chrysalis. 



In a communication read before the Academic des Sciences 

 (at the meeting of June 28th, 1897), I pointed out in the 

 Phasmidai a sensible ditference between the rate of growth of 

 a normal limb and one in process of regeneration, which was 

 to the advantage of the latter. This difference is even more 

 noticeable in the Blattid;^, and greater still in the Mantidaj. 

 In the last-mentioned Orthoptera, when self-mutilation has 



* Limbs in process of refreneration develop in the snme way among 

 the Orthoptera saltrttoria. Such a process is also constant in regenera- 

 tion that follows artificial removal <ii" the tarsus and the distal end of the 

 tibiii, althdujrh at tirst sijrlit certain dift' rences may apiiear to exist. The 

 rule oufrht prohably to be a i-eueral one in the Arthropoda, in which 

 reiieueration of l^st appendaj^es (includinfr auteuuie) takes place. I am 

 surjjrisetl that M.s>rs. Halexm and Hriudley have not recorded thia 

 remarkable point in their line work on the Blattidae. 



