Echinoderm LarvoB ivith Two Water- Vascular Systems. 327 



is then passed through a Berkfeld filter in order to remove from it all 

 bacteria. 



The ripe eggs obtained from the females were fertilised in finger bowls, 

 and the next morning, when the free-swimming blastulae had risen to the 

 top these were decanted off into " Brefifit jars" in which a considerable portion 

 of, and in some cases all, their further development took place. 



These " Brefifit jars " — the use of which was introduced by Dr. Allen and 

 the workers in the Plymouth Marine Biological Laboratory — deserve some 

 words of comment. They are made of thick green glass, and each contains, 

 when fuU, about 2^ litres of sea-water. The colour of the glass, which 

 admits approximately only the same rays of Light as can penetrate the 

 surface-layers of the sea, seems to be of some importance in keeping the 

 larv£E healthy. The jars were placed in a sink which was kept moist by a 

 trickle of water from the tap. 



The larva is capable of developing for four or five days without food, 

 indeed the ectodermal oesophagus only joins the endodermal stomach on the 

 fourth day (fig. 1). By the fifth or sixth day the initial impetus of develop- 

 ment has been exhausted, and although the larva can live for several days 

 longer, degenerative processes set in, and sooner or later it falls to the bottom 

 and dies. The degeneration shows itself especially in the absorption of the 

 ciliated arms. 



It is therefore necessary to supply the larva with food, and the food chosen 

 was a pure culture of the diatom NitzscMa. These pure cultures are pre- 

 pared in the Plymouth Biological Station, and directions for sub-culturing 

 are given in Dr. Allen's paper on the subject (1). 



The successful rearing of the larvae depends principally ou the establish- 

 ment of a healthy growth of NitzscMa in the culture jar ; but according to 

 Dr. Allen's experience, not more than about thirty larvse will grow to an 

 advanced stage of development in a Brefifit jar, and of these ordinarily only 

 a few will complete their metamorphosis. It is, however, possible to rear 

 hundreds of larvae in such a jar until they are about a fortnight old. But 

 for a successful issue of a rearing experiment, it is necessary as the larvae 

 grow older to thin out the culture at frequent intervals by transferring 

 portions to fresh Breffit jars, and niy experience coincides with that of 

 Dr. Allen that one is lucky if one succeeds in rearing six or seven larvae in 

 each jar through the metamorphosis. 



In 1914, 1915, and 1916 I reared the larvae in Breffit jars, but I was 

 dissatisfied with the numerical results obtained, and in 1917 I adopted the 

 plan of transferring the larvae when they were about ten days old, into much 

 larger receptacles, viz., the plunger jars. 



