Delsman : Fish Eggs and Larvae. 



39 



A third peculiarity is the segmented yolk, so characteristic for the eggs of 

 herring- and eel-hke fishes, and which always gives a strong indication as to 

 the direction into which we have to look for the origin of any pelagic ^gg. 



Finally a num- 



^Jr%^, 





Fig. 2. Part of the egg-membrane stronger enlarged (440 X), 



ber of small oil- 

 globules distributed 

 irregularly in the 

 yolk make this tgg 

 one of the most 

 easily recognizable 

 among the numer- 

 ous kinds occurring 

 in the Java Sea. 



In different eggs I counted from 7 to 19 of these oil-globules, the number 

 being very variable. 



I can not say how long the period of incubation is. On several occa- 

 sions I have collected a considerable number of these eggs from the horizontal 

 surface catches mentioned in the first of this series of articles, but they 

 all showed the rudiment of the embryo already, so that I could not guess 

 their age. Probably, however, the incubation will not take more than 1^2 

 or 2 days. The time of hatching was always the same, viz. between 8 and 

 9 o'clock in the evening of the day they had been caught, though sometimes 

 a few might hatch a little earlier or a little later. From this circumstance 

 we may conclude that there is also a fixed spawning time, although I could 

 not make out at which time of the day this is. 



Fig. 3. Newly hatched larva, enlargement as fig. 1. ot. ear-vesicle, a. anus. 



An examination of the newly hatched larva confirms at once our surmise 

 that we are in this case dealing with an ç:gg belonging to a fish related 

 to the herrings. The backward situation of the anus, together with the general 

 appearance of the larva and the segmented yolk, put this beyond doubt. 

 The postanal part of the body, as shown in fig. 3, is less than V7 of the 

 total length though, in somewhat older larvae, this proportion gradually 

 changes in favour of the tail, being 1 to 5V2 e.g. in fig. 4. The head is 



