MUSEUM OF COMPARATIVE ZOOLOGY. 155 



dreae ('81, p. 220) ! The secretion may also be of use in cementing 

 the sand grains into a sort of tube noticeable when the animals are dug 

 out of the sand. 1 



The multicellular glands present a type easily distinguishable from that 

 just described. They are to be met with everywhere, not only in the 

 papillae, but lying in the interspaces as well, and extending up into the 

 clear zone of the introvert, where they are the only differentiated hypo- 

 dermal cells. Never much crowded, they become here sparser, until they 

 completely disappear at the level of the upper papillate zone ; nor are 

 they to be found in or above this zone, nor at any point on either sur- 

 face of the tentacular fold. The multicellular glands may be identified 

 on surface preparations, but an insight into the histological relations is 

 first afforded by sections. With hsematoxylin the cell body stains lightly 

 but uniformly, the mass at the distal end more deeply (Fig. 12), but 

 with this stain no nuclei can be found either in the cells or in the con- 

 nective-tissue investment of the gland. Each gland is seen to be made 

 up of a number of flask-shaped cells, which are separated by thin par- 

 titions and which unite at their distal ends into a duct piercing the 

 cuticula and opening upon its surface to the exterior. Andreae ('81, 

 p. 216) was unable to find any nuclei in these cells. The application 

 of a carmine stain, however, shows their presence near the proximal ends 

 of the cells (Fig. 14), where they often lie flattened against the cell 

 membrane by the crowding of the granules accumulated in the cell 

 plasma. The same stain demonstrates also (Fig. 13) smaller nuclei at 

 various points in the connective-tissue investment. There is likewise 

 seen to be a difference in the cells of any one gland which indicates 

 alternation in secretive activity. Thus the plasma of some cells is 

 thickly crowded -with large granules, which are entirely wanting in other 

 cells. This is most clearly demonstrated in a transverse section of the 

 gland, as shown in Figure 13. The cells differ in intensity of color to 

 correspond with the number of granules present, and large distended cells 

 are found near those which are evidently thinner and poorer. The pro- 

 duct of these glands is a substance more waxy than fluid, to judge from 

 its manner of caking in the duct, and breaking up into small fragments, 

 like sebaceous material. Tts discharge is evidently gradual like its pro- 

 duction ; for I have never found a gland empty, nor does the total 

 amount of secretion present vary greatly. This alternation in func- 

 tional activity between the various cells of one gland and the constancy 



1 For this suggestion, and the observation that such a tube exists, at least for 

 S. Gouldii, I am indebted to my friend, Mr. C. B Davenport. 



