242 Dr. J. D. F. Glldnist 07i 



the cilia, and was always in the direction of the long axis of 

 the body, towards the thicker end. At no time were they 

 seen to j)rogress in mid-water; when dropped into the water 

 they sank to the bottom and commenced crawling about. 

 The rate of motion was often fairly rapid — an inch in seven 

 seconds in one case — and the larva readily surmounted any 

 obstacles in its patli by crawding up and over them. They 

 were slightly adherent to the substance over which they 

 moved, so that the dish which contained them could be tilted 

 under the microscope without altering the position of the 

 larvaj by the movement of the water over them, except 

 when this was continued for some time. Whether any 

 mucus or adherent substance was secreted by the lower 

 surface of the larva when in motion was not observed, but 

 this seems Jiot improbable, as sections reveal a thick foot-like 

 area of epithelial cells on the anterior half. On one occasion 

 a specimen picked up in a pipette became so firmly adherent 

 to the glass that it was with difficulty removed, and it was 

 observed to adhere most Hrndy by one end, the other moving 

 freely with the passing of the watcir up and down the tube. 



Suvxmary of Results. 



1. The normal habitat of Cephalodiscus in S. Africa 



appears to be rocky ground in fairly shallow water, 

 but bidow low-water mark. It is either attached to 

 rock or some substance growing on rock, but may 

 become detached and carried on to muddy ground. It 

 seems to be abundant on the south coast. 



2. It may grow from a small basis, or the basis may be a 



broad sheet of ccenoecial substance from which several 

 main stems arise. 



3. In the living state the zooids and buds have been ob- 



served on the general surface of the coencecium outside 

 the tube, and sometimes at a distance from it. 



4. The buds in such cases act as anchors, being firmly 



adherent to the surface by their proboscides. 



5. In both bud and zooid a quantity of viscid mucus occurs 



between the proboscis and the coencecium. 



6. There is no evidence that the buds ever develop into 



normal zooids, and they may be individuals specialized 

 for adhesive purposes and coenoecium-building. 



7. More than one zooid and its buds may occur in one tube 



in the coencecium in the Cape species. 



8. Buds and zooids are provided witli cilia over the whole 



of their surface. Their stolons or stalks are also 

 ciliated. 



