156 



Part III. — Ticenty -first Annual Heport 



The continuity between the cavity of the limb and that of the gill is 

 well seen in this stage. 



III. (Fig. 74). The gill of the chela is now double. 



IV. (Figs. 67 and 79). The separation of the limb into its basal part 

 (pleuron, ap) and the limb proper has taken place, and the gills of the 

 pereiopods are, although not yet functional, now large, and show their 

 relation to the limb proper. 



The epipodite and two gills of the third maxillipede are well 

 developed. 



The process on the second maxillipede which was noticed on the 

 I. Zoea is now prominent (Figs. 66, 67, and 174). It represents 

 probably the epipodite and one gill. 



Megalops (Figs. 80, 64. and 86).— The gills connected with the 

 pereiopods are now functional, and they occupy the positions they do in 

 the adult. The two arthrobranchs of the chela are attached to a space 

 between the pleuron and the coxopodite, ar-b?' {'2); in the case of the 

 second and third pereiopods the pleurobranch is attached to a cii'cular 

 opening in the pleuron (pl-hr, Fig. 80.) 



On the second maxillipede (Fig. 64) the gill and epipodite are attached 

 to the first joint of the protopodite. They both rise from the same 

 stem. The gill is a podobranch, and is yet unsegmented. The second 

 gill, viz. the arthrobranch, which is found attached to this appendage in 

 the adult was, if present, not noticed. 



On the third maxillipede (Fig. 86) three gills are found in this stage, 

 none of which are segmented. Two are arthrobranchs ; they are the 

 gills present in the Zoea stages. There is also a small gill attached to 

 the second protopodite joint close to the origin of the epipodite ; it and 

 the epipodite are probably developed from the long epipodite process 

 seen in the Zoea. 



1. The podobranch of the second maxillipede {br, Fig. 71) shows 

 some segmentation in its proximal part. A second plain gill (br ?) was 

 made out in one case. The gills are very liable to be knocked off in 

 dissection, and unless the maxillipedes are separated from one another 

 the relations of the parts are not satisfactorily made out. 



On the thii'd maxillipede (Fig. 88) only one of the gills is divided into 

 lobes — viz., the larger arthrobranch. The other arthi'obranch and the 

 little podobranch have still the smooth outline. 



In my paper on the larval stages of Crangon vulgaris it was stated, 

 with reference to the gills of the pereiopods, that starting from the 

 podobranch condition they passed thiough an arthrobranch (non- 

 functional) stage (V.) and that in the Megalops (YI. stage) they appear 

 as functional gills (pleurobranchs) for the fii'st time. Re-examination 

 of these stages has show^n that the interpretation of the developing gill 

 in the V. stage as an arthrobranch is wrong. In that stage it is a 

 pleui^obranch. 



The description given above of the epipodites and gills shows a close 

 similarity between them in their origin and development. This is 

 especially the case in the second and third maxillipedes, where a 

 podobranch and an epipodite are apparently derived from the one bud. 



The development of the epipodite of the first maxillipede is included 

 in the description of that appendage, p. 148. 



[Table. 



