MATERIALS FOR A MONOGRAPH OF THE ASCONS. 3819 
placed very close to the gastral epithelium (figs. 50, 52). 
Examination of preparations not manipulated in any way 
shows that the founder, as in the case of the monaxons, does 
not leave the ray until it has attained its full length (figs. 29, 
97,99). In C. clathrus, as I showed formerly, the founder 
does-not leave the ray at all and is thus responsible for the 
clubbed ends of the rays which are distinctive of this species. 
As soon as the ray has grown to its full length, however, 
the founder disappears. The thickener has, meanwhile, 
migrated slowly outwards from the base of the ray, depositing 
lime and finishing off the ray as it passes along. Shortly 
before the disappearance of the founder, the thickener is 
found close to it just as in the monaxon spicules (compare 
figs. 97 and 99 with figs. 7,8, 17, 18, 57, 7la, 82, 83, 85—88). 
In the fully-formed triradiate each ray bears the thickener 
at the extreme tip, generally in the form of a compact cell, 
but sometimes remarkable for sending out slender processes 
in various directions (figs. 97, 98). The thickener does not 
quit the ray, but can always be found attached to the tip of 
it. Woodland, however, states (1905, p. 145) that in Sycons 
both formative cells quit the rays of the triradiates. I can 
only say that in all the Ascons I have examined, I have never 
failed to find at least one cell on the apex of each ray of the 
full-crown triradiate systems. 
IV. THe Gastrat Rays. 
In the genus Leucosolenia the gastral rays of the quadri- 
radiates are always curved and thorn-like, the apex pointing, 
typically, towards the osculum, i. e., in the direction of the 
water-current. , 
In the formation of the gastral rays, I found a curious 
difference between those formed in the oscular rim, above the 
level reached by the gastral epithelium of collar-cells, and 
those formed throughout the gastral cavity generally where 
it is lined by collar-cells. I will deal with the latter first. As 
