940 MR. H. H. BlUXULEY OS REPRODUCED [DeC. 13 



" normal " tarsi. The interest of these crippled tarsi lies in their 

 great rarity, their tendency to possess several supernumerary 

 spines, and the fact that the fusions resulting from the incom- 

 pleteness of the articulations are nearly always confined to the 

 intermediate shorter joints, leaving the proximal and distal joints 

 \vell-defined. 



The Characters of REPRonucED Appendages in other 

 Insecta and Arthropoda generally. 



Before making a comparison of the normal and reproduced legs 

 of the Blattidse from a statistical point of view, it is desirable to 

 refer to certain special characters possessed by reproduced appen- 

 dages in other groups of Arthropods. 



The observations of various authors on reproduction of lost 

 appendages in Arthropoda have usually been of but partial nature, 

 and only in a few cases have experiments in this connection been 

 extended over a large number of individuals of the same species. 



But the published work is sufficient to indicate that the 

 structures which replace lost or mutilated appendages fall into 

 two main divisions. Moreover, one or the other of these two 

 kinds of reproduction is constantly associated with a particular 

 degree of injury in particular cases of appendages or genera to 

 the entire exclusion of the other kind. In other words, a certain 

 injury to a particular appendage among particular families of 

 genera is invariably followed by reproduction of one kind. The 

 other kind of reproduction is as constantly associated with other 

 cases. The two kinds of reproduction met with are briefly : — 



(a) In all chief respects, such as the numher of joints and their 

 ■relative dimensions, the reproduced appendage is the counterpart of 

 the normal congenital appendage. 



(b) The reproduced appendage differs frotn the normal appendage 

 in certain respects which are constant, and in cases where maturity of 

 the animal is attained through a series of ecdyses the special features 

 of the reproduced appendage are perpetuated ; so that, strictly 

 speaking, the animal does not reproduce the normal appendage. The 

 chief di'itinguishing feature of this hind of reproduction is that the 

 numher of joints present is less than in the normal appendage. 



The roLLOwiNG cases fall under (a) : — 



Crustacea. Decapoda. — Observations on the reproduction of 

 lost appendages appear to have been mainly on the Decapoda. The 

 accounts of authors already referred to agree in stating that the 

 chela; and u<alking-legs are reproduced by structures resembling the 

 normal in all respects except size. As regards the flagellum of 

 the antenna, it is, however, not possible to speak with certainty, 

 for no observer has given any details as to the number of its joints 

 when reproduced. Moreover, the large number of joints of the 

 normal antenna in most cases and the liability to loss of the more 

 distal joints leave the normal numbers for most species somewhat 



