470 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxix 



(3) In all directions from the yolk, the streaming goes to the ger- 

 minal disk {Oara-sslus^ Kowalewski, ls86). 



As best I can determine, the pipefish comes under class two. This 

 matter will be further referred to in the section dealing with the 

 periblast. 



Intimately connected with the foregoing is the collecting of the oil 

 drops underneath the germ disk. In pelagic eggs, generally the oil is 

 in one great globule near the center of the yolk, l)ut in the pipefish 

 many small orange-red globules are imbedded in the periphery of the 

 yolk. When the protoplasm moves up to the animal pole, the oil 

 globules go also and are collected under the germ disk to form the 

 '"'' disque huileux'''' of Lereboullet. This is a phenomenon very common 

 among Teleosts. It has been reported b}^ all workers on the Salmo- 

 noids, by Ransom (1867) for the stickleback, Kowalewski (1886) for 

 Oarassius, and by many others. Rathke (1837) tirst described these 

 processes in pipefishes from the Black Sea. He says that the ger- 

 minal disk is formed after the eggs come into water, and that the j^ellow- 

 red '"^fetV drops which surround the yolk flow up to and spread out 

 under the disk in a layer covering about one-third of its upper surface. 

 Kupfier (1868), describing the ^.gg of a European form, says, "This fat 

 forms a mass of drops of difierent sizes, which incloses the germ disk 

 underneath and laterally." 



The two phenomena described above are intimately connected with 

 and in fact bring about another known as the " clearing of the ^gg^ 

 As the protoplasm is withdrawn from the center and the oil globules 

 from the periphery, the pipefish ^gg becomes "clear;" that is, the 

 yolk, freed from these substances, becomes homogeneous and trans- 

 lucent. At this stage the ^gg of Sijjhostoma (Plate V, fig. 1) consists 

 of a button-shaped protoplasmic disk resting on an orange-red layer 

 of oil globules embedded in yolk and covering about one-fourth of the 

 ^ggi the other three-fourths consisting of clear milky yolk. This 

 " clearing" has been described, essentially as above, by Fusari (1890), 

 Kowalewski (1886), and Agassiz and Whitman (1885), for Cristiceps^ 

 Carasslus^ and Ctenolahriis^ respectively. 



In connection with the above processes, many workers, especiall}^ 

 the students of the Salmonoids, have described amoeboid movements 

 of the germ disk, and His, in a recent paper (1899), has described such 

 activities in the blastomeres up to the sixteen-celled stage. Ransom 

 (1867) has also figured and described amceboid movements in the yolk 

 of Gaster'osteus. These movements seem to assist in freeing the yolk 

 of protoplasm and the germinal disk of yolk. The opacity of the e,gg^ 

 which prevented my making out much about the " streaming," oper- 

 ated here against the detection of such movements. Once or twice, 

 however, I thought that I did make them out, and in several hardened 

 germs there were found such protuberances as are figured by Henne- 

 guy (18SS) in trout germs hardened in chromic acid. 



