186 
W. BALDWIN SPENCER. 
The only change in their position is produced by the deve- 
lopment of reproductive products which lie between the tubes 
and the alimentary canal, and from which, though in very close 
contact, the former may be seen to be separated by a distinct 
membrane. Fig. 5 shows diagrammatically the relative positions 
of the organs of Gammarus pulex. 
These organs in the Gammaridse are treated of in considerable 
detail by Nebeski 1 in his account of the Amphipoda, and he 
describes them as being present in very varying stages of deve- 
lopment in different members. In Melita a single and very 
rudimentary one is present. Corophiiden and Mcera each 
possess a pair of small ones, whilst in Gammarus and Cyrto- 
phium the pair are much better developed. In Orchestia not 
only were the tubes still more prominent, but he found within 
them what he states are excretory products in the form of con- 
cretionary bodies, each of which takes its rise within an epi- 
thelium cell of the tube, and, gradually growing, pushes the 
cell substance aside and comes to lie within the lumen. These 
bodies, which appear from his description to exactly resemble 
in form others which I have found in the closely allied Tali- 
trus locust a, Nebeski states consists of calcium carbonate. 
Gamroth 2 has described the presence of the tubes in the Ca- 
prellidae,and states that though their intimate structure and phy- 
siological import is unknown to him, yet he has found granular 
concretions in them, and regards them as excretory organs. 
When transverse sections of the hinder part of Caprella are 
cut their existence is very easily recognised, as they form very 
prominent pouches lying one on either side of the walls of the 
alimentary canal communicating with the latter by somewhat 
constricted openings. Nothing apparently enters these sacs 
(for such they are in form in Caprellidse rather than tubes) from 
the gut, as in sections in which the latter is full of food the 
sacs themselves are quite empty. Their walls consist of elon- 
gate nucleated cells, just as in Gammarus and Talitrus. 
1 4 Beitrag zur Kentniss der Ampliipoden der Adria,’ Zoo). Institute, Wien 
Band iii, 1881. 
2 ‘ Zeitscbr. f. Wiss. Zool.,’ Bd. xxxi, p. 115. 
