136 Transactions of the Society, 



three categories of substances, it is possible that any one of the 

 categories may be either absent or reduced to a very small quantity. 

 Thus the yolk of the sponge Qgg seems to contain a very slight 

 amount of protein, as is indicated by fixing and staining tests. As 

 has already been mentioned, the percentage and kind of fat may 

 vary also, even among closely allied forms. Many workers now 

 consider that the mitochondria are formed of a protein substance 

 associated with a lipin, and microchemical tests appear to bear out 

 this assumption. The constitution of the Golgi apparatus is pro- 

 bably much like that of the mitochondria — i.e. proteid in some way 

 linked with lipin materials. 



Any collection of granules or discs within the Qgg cytoplasm 

 which behaves according to a majority of these tests may be called 

 "yolk":— 



1. Granules which do not go red in Altmann, Champy-Kull or 

 Bensley-Cowdry (i.e. granules which are not fuchsinophile). In 

 the case of Champy-Kull, grains which go blue, or in the case of 

 Bensley-Cowdry grains which go green, are likely to be chromatin. 

 (Insect yolk, however, may be fuchsinophile.) 



2. Granules which are yellowish, brown or black in Champy- 

 Kull, Benda, or chrome- osmium fixation, i.e. Flemming or Her- 

 mann's fluids. 



3. Granules which do not go black after staining in iron- 

 haematoxylin, following two days' fixation in neutral (5 p.c. to 

 20 p.c) formalin. (There are notable exceptions to this, e.g. the 

 yolk of Amphibia, which stains intensely in iron-hsematoxylin.) 



4. Granules which do not disappear after fixation in corrosive 

 acetic acid, Bouin, Carnoy or Petrunkewitsch. (These are many' 

 exceptions to this, e.g. the delicate yolk of sponge, and some 

 mollusc eggs.) 



5. Granules, which rarely go black in Mann-Kopsch or Kopsch 

 (but which, if they do go black, are easily decolorized by a short 

 immersion in turpentine, and are then left as yellowish spheres, 

 and not as vacuoles) ; granules which will not stain red after such 

 extraction of their colour in turpentine. 



6. Granules which are greyish or yellowish intra vitam (in 

 some cases, however, the mitochondria may be loaded with a 

 yellow pigment). 



7. Granules which do not go black in Cajal's silver impregna- 

 tion method. 



8. Granules which exceed 2//, in diameter. 



Yolk-Formation in the Sponge Grantia. 



Eecently the oogenesis of Grantia comprcssa was studied by 

 one of us (19), At no period of oogenesis will the usual techniques, 



