'I'HE CCELOM AND NEPHRIDIA OF PALiEMUN SELUiATUS. 167 



7 tlie entry of three such tubules, a, j3, and y, is represented. 

 The end-sac itself is a kidney-shaped structure, receiving in its 

 concavity a large blood-vessel. It contains in its interior a consider- 

 able cavity, into which project a number of radial septa. These 

 septa, together with the external wall of the organ, are made up of a 

 dense connective tissue, which is deeply stained by ha3matoxylin, and 

 in which run numbers of blood-vessels (fig. 7, b. v.). These blood- 

 spaces are not, however, so numerous as those figured by Grobben in 

 the corresponding septa of P. Treillianus ; and there is a further 

 difference between the blood-spaces of Grrobben's figure and those 

 seen by myself, inasmuch as the latter are always bounded by a distinct 

 epithelium. I do not wish, however, to throw doubt on the accuracy 

 of Grobben's figure without having had an opportunity of investi- 

 gating the species described by bim. 



The cavity of the end-sac is lined by a curious epitbelium, com- 

 posed of large, pale, finely granular cells, with rounded nuclei (see 

 fig. 7, Ge. ep.). In none of my preparations have I been able to see 

 an arrangement of the epithelium so regular as that figured by Grobben 

 (1. c, pi. i, fig. 8). In many of my sections, also, patches occur on 

 which the epithelium is absent, — apparently from some cause inde- 

 pendent of the methods of manipulation. 



The cavity of tbe " end-sac," and to some extent also the lumen 

 of the nepbridial tubes, is filled with an irregular, finely granular 

 clot. For the sake of clearness, this clot has only been inserted in 

 the upper half of fig. 7. 



The arrangement, of which a description has here been attempted, 

 will be made clear by an inspection of the diagram fig. 6, in which 

 it will be evident tliat the comparison so often made (by Glaus, 

 Grobben, and others) between the glomerulus of the Vertebrate 

 kidney and the end-sac of the Crustacean green gland is abundantly 

 justified, each glomerulus being the termination of a caecal outgrowth 

 from a bent nephridial tube, which communicates on the one hand 

 with the body-cavity and on the other either directly or indirectly 

 with the exterior. 



In a future paper I hope to describe the modifications of this ar- 

 rangement in other genera. I may here say that I have found in 

 Pandalus annulicornis a coelomic sac which is similar in its relations 

 to the sac just described, but that in this genus the glandular part 

 of the kidney is formed entirely by the glomerulus, — this structure 

 being formed by a folded invagination of the wall of the nephridial 

 tube, containing large numbers of blood-capillaries. 



iiephridium is established by more than one tube, occurs in the young coxal gland of 

 Limulus (cf. Gulliuid, Quart. Journ. Micr. Sci., vol. xxv, pi. xxxvi, fig. 2). 



