BTOLOGY. 
Spang. 9 
d. Epithelia and the Cuticula ( 5 , 61 ). 
Bidder (5) has very carefully examined the collar-cells of Grcintia 
compnssa and other forms, with a view to find fresh proofs for the posi- 
tive statements previously made by him concerning the existence and im- 
portance of the membrane connecting the collar-margins of these cells, 
known under the name of Sollas’s membrane. The result of his work was a 
negative one ; he was obliged to corroborate the statement made first by 
the Recorder (in 1889), and later by Yosmaer & Pekelharing (in 1893), 
th it in life no such membrane exists, and that, where it is found, it is a 
post-mortem artefact, produced by want of care in hardening. 
In Grantia compressa the collar is made up of about thirty parallel 
rods, united by a film of some other substance. The flagellum is inti- 
mately connected with the nuclear membrane. Concerning the existence 
of a substance sometimes occurring between the collar-cells, the disco- 
very made previously (in 1889) by the Recorder is here for the first time 
corroborated. The smallest collar-cells were observed by him in a Sponge 
i lentical with, or at least very similar to, Acanthella stipitata , Carter. 
Here they measure only 0-0017 mm. in height, and 0-0008 mm. in basal 
width. Among the largest collar-cells observed were those of Grantia 
compressa , which measure in life 0 012 mm. in height, and 0-0066 in basal 
width. 
Topsent (01) corroborates the statement previously made by the 
Recorder, that in Corticium candelabrum the canals are clothed with a 
high epithelium, the elemonts of which are cylindrical, and have a convex 
free terminal faco, which is destitute of a flagellum. On the other hand 
flagellated epithelial cells were found by him in the walls of the wide 
canals of Placina , in Oscarella, Placortis simplex , and Placinastrella 
copiosa. In Placina these flagella are stouter than the flagella of the 
collar -cells ; in Oscarella they are particularly long. By the possession 
of these flagellated epithelial cells these Sponges are distinguished from 
the other groups in which no such flagellated epithelia occur. He has 
found a well-developed cuticula on the outer surface of Corticium cande- 
labrum. This is sometimes detached in the sections. He gives a photo- 
graph of such a section with a partially detached cuticula 
e. Cortex ( 61 ). 
Topsent (61) finds the cortex very variously developed in the Carnosa. 
In JDercitus a thin ectochrote with microhabds and a collenchymatous 
layer, with large spherical cells, are met with. The latter contain — in 
D. buclclandi — pigment granules. In D. plicatus these elements are par- 
ticularly large, twice the diameter of the homologous cells in Ancorina 
mucronata. In the living state they are filled with yellow, non-refringent 
globules. In paraffin sections they appear as vesicles, the contonts of 
which are amassed in the centre, these forming a sort of nucleus. In 
the Microsclerophora the cortex is chondrenchymatous, that is to say, 
composed of a substance “ fondamentale anhiste,” which contains spherical 
