154 NOTES ON THE REPRODUCTION OF TELEOSTEAN FISHES 



of the tail. A few occur on the ventral part of the trunk and on 

 the rectum. There are several large dendritic black chroraatophores on 

 the dorsal near its margin, while two are seen on the post-anal part of 

 the ventral. A patch of pigment occurs above and below the caudal 

 extremity. In all cases the black pigment of the marginal fins is 

 accompanied by paler chromatophores, the colour of which has been 

 destroyed by the reagent. I cannot speak definitely as to the nature of 

 the notochord. The stage of development suggests most strongly that 

 the larva is derived from a pelagic egg. It may possibly be a belated 

 specimen of Callionymus lyra, but it does not closely agree with any 

 examj)le of that species which I have seen. 



Clupea harengus. — Linn. Herring. 



The young stages of the herring may be most conveniently reserved 

 for consideration in connection with the distribution of young fishes in 

 the Plymouth district. T only purpose at present to call attention 

 to the occasional occurrence of an abnormal feature in the egg. It 

 is well known that the ova of the pilchard and sardine and of the 

 shads are characterised by the formation, after immersion in sea water, 

 of a very large perivitelline space. The sprat ovum has only a very 

 small perivitelline space, while that of the herring ovum is normally of 

 moderate proportions. It is difficult to take accurate measurements if 

 the spawn is allowed to adhere together, since the zona is then apt to 

 assume a polyhedral form, but this can be obviated by the use of starch 

 as recommended in the United States Manual of Fiah Culture. I find 

 among the spawn of three fish treated in this way, that the largest normal 

 eggs measure TTG mm. in total diameter, the yolk mass measuring 

 1*25 mm., 24 hours after fertilisation. There are, however, several eggs 

 of a much larger size. One slightly elliptical, but not at all flattened, 

 has the greatest diameter 2'42 mm., the least 2-34 mm. ; it appears to be 

 as large as any. It must be remarked that these specimens are all 

 dead, as in all previous instances of abnormally large herring eggs 

 which have come under my notice. So far as I can judge, the excess in 

 size is confined to the perivitelline space, but it is not the case that 

 dead herring ova have usually a larger space, that is to say a more 

 inflated zona, than living specimens. The latter are usually the 

 larger, at least when death takes place at an early stage. 



I do not suppose that this observation is new to those who have had 

 to deal with herring spawn, but I do not remember to have seen it 

 recorded. It shows that the perivitelline space may exhibit excep- 

 tionally an approach to the dimensions normal in the shads and 



