460 INVEETEBEATA chap. 



results may be attained by the following method :— The larvae are 

 deluged with the strongest formalin (40 per cent solution) ; this is 

 neutralized before being used by having a piece of chalk immersed 

 in it. The formalin is only allowed to act for a couple of minutes, 

 and the larvae are at once transferred to absolute alcohol, to which 

 a drop of strong ammonia has been added. They are then stained 

 in a solution of either eosin or safranin in absolute alcohol, which 

 must be allowed to act for at least twenty- four hours, preferably 

 for several days. Then, drop by drop, at considerable intervals, oil 

 of cloves is added, — it is best to prolong the period of the addition 

 of oil of cloves over several days, — and then the larvae are transferred 

 to pure oil of cloves for several days. Finally, they are placed in 

 the well of a concave slide and a drop of a thick solution of Canada 

 balsam in xylol is placed upon them ; the oil of cloves flies off by 

 surface tension to the periphery, and can be — if the operation is 

 skilfully performed — almost entirely removed by blotting-paper. A 

 coverslip is now gradually pushed over the preparation from the 

 side, so as to avoid the formation of bubbles, and in this way a 

 permanent preparation is made. 



For younger stages, where there is not so much gelatinous tissue 

 as in the older larvae, a simpler method has been invented by Professor 

 Graham Kerr. He iixes them by immersing them (when freed from 

 as much of the salt water which clings to them as possible) in 

 absolute alcohol. They are stained in safranin dissolved in absolute 

 alcohol, and are then mounted in halsain dissolved in ahsolute alcohol. 

 Only a thin solution of this can be obtained, and it shrinks greatly 

 in the drying, but by patient addition of fresh solution, as that which 

 is just added dries, very beautiful permanent mounts may be obtained. 



ASTERIAS 



The egg of Asterias segments with great regularity into blasto- 

 meres of approximately equal size. It is a beautiful example of 

 indeterminate segmentation. The result of segmentation is a hollow 

 blastula, and it can be shown experimentally that, up to the 500-cell 

 stage, this blastula is not functionally specialized in any way, but 

 that any sufficiently large fragment cut from it will heal up by the 

 approximation of its edges, and so form a miniature blastula which 

 will develop into a perfect miniature larva. At about the 1000-cell 

 stage the cells develop cilia, and the blastula begins to rotate within 

 the egg-membrane, which it soon bursts, and it then rises to the 

 surface of the water and begins its existence as a free-swimming 

 larva. Certain of the Porifera and of the Coelenterata are the only 

 species of animals, outside the phylum Echinodermata, in which the 

 larval existence is begun as early as it is in the development of 

 Astei'ias. 



At the end of a day of free-swimming life the blastula begins to 

 be converted into a gastrula. The blastula loses its spherical shape 



