270 BULLETIN OF THE BUREAU OF FISHERIES. 



molting. " It is to be further noted that as early as the third larval stage and for some 

 time thereafter the claw-tip, like the tooth, gives passage to the duct of a gland (d. t. g., 

 fig. 19-20). I have not found glands of this type in the spines of the adult claw, and 

 if present in older adolescent lobsters they are successfully concealed by the opacity 

 of the shell. The adult spines were sectioned, but in all the young stages glycerine 

 preparations were relied upon. A single tooth sometimes bears the ducts of three inde- 

 pendent glands, in which case it is probably compound, resulting from the fusion of a 

 corresponding number of teeth. Rarely a bifurcated duct is seen (fig. 2 pi. xlii), each 



Fig. 21 and 22. — Right and left forceps of lobster 24 mm. long, reared in captivity, and 11 months old, in 

 eighth or ninth stage; seen from above, showing early state in the differentiation of cracker (right) and 

 toothed claws. Enlarged about 40 times. L, lock spine, as in all figures. 



tube issuing from a separate gland, but with common opening at the summit of tooth. 

 Whether these organs possess any special significance in these parts or not I am unable 

 to say. 



The first step in the differentiation of the cracker claw, as already remarked, is 

 seen in the rounding or blunting of the teeth, particularly at the proximal end of the 

 series (see fig. 22 and 24, and especially fig. 25). The teeth appear to be retarded in 

 growth, and while these remain blunt and irregular, those of the toothed claw become 



o The sensory hairs, as already stated, are derived solely from the epidermis, no mesoblast ever entering them, and they are 

 invaginated with every molt. The claw teeth are tubular outgrowths of the wall of the appendage, and are never invaginated. 

 The rostnmi, as well as at least the tips and terminal spur or tubercle of the propodus, are seen to arise like the setae, and like them 

 are invaginated during the early molting periods, but they are eventually entered by mesoblast. 



