EXPERIMENTS ON ASYMMETRICAL FORMS 257 



produced by regeneration, because the structure of its tubercles 

 and teeth does not correspond to that of a large crusher claw, but 

 of the nipper claw, while the nipper on the left side has not yet 

 been transformed into a crusher. We learn from this abnormal 

 case that the right chela is not only capable of taking on characters 

 of the normally small, left claw but also of developing nipper dac- 

 tylopodites with the symmetry of both the right and left sides. 

 But the secondary homochely in this case prevents us from solv- 

 ing the problem, which was presented at the beginning of the 

 paper. We must look for an object, in which the heterochely 

 remains unchanged when a supernumerary appendage with 

 secondary symmetry has arisen. The lobster furnishes such 

 cases (Homarus, fig. 7) because if the well developed crusher 

 is removed no reversal takes place, but a new crusher regenerates 

 in the old place, the nipper claw remaining unchanged. But 

 during their regeneration the large claws pass through stages 

 which resemble the nipper claw, having an equal, fine dentition, 

 interrupted by enlarged teeth, while the true crusher is character- 

 ized by a rough and irregular dentition. 



Monstrosities with triple chelae are not infrequently met with 

 in the lobster and are undoubtedly due to regeneration, because 

 such have also arisen in experiments, and furthermore because 

 abnormal appendages, very apt to be thrown off during the 

 successive moults, are generally found in lobsters of considerable 

 size. In the course of the past few years I have been searching 

 through the literature for references to such monstrosities, and 

 have especially searched through museum-collections and have 

 made a comparison of these triple chelae of lobsters, the results 

 of which may be summarized thus: 



1. Supernumerary appendages with secondary symmetry in 

 the small or nipper claw exhibit always the characters of nippers 

 and even those with the symmetry of the opposite body-side 

 form no part of the large, or crusher-claw, as will be observed in 

 the photographs, figs. 8 and 9. 



2. Supernumerary appendages with secondary symmetry in 

 the large, or crusher claw, have the same characters in both divi- 



