sperm of any of the starfish. However, by changing the constitution of the sea 
water he succeeded in getting impregnations (in some cases 50 per cent.) 
between S. purpuratus 9 and Asterias ochracea, Segmentation was normal ; 
the laryze developed into blastulee and gastrule, some showing the differ- 
entiation of the intestine. Many other experiments in hybridizing fishes 
have been recorded. These, however, were all between nearly related 
species, mostly among the domesticated salmonidée and cyprinide. It would 
not be to the point to pass these in review here. For a good summary of 
these the reader is referred to Ackermann (’98). 
METHODS. 
The method of effecting the crosses and the precautions taken to pre- 
vent contamination with other sperms, were in all cases essentially the 
same. The sexes of the same species were kept in separate aquaria. The 
eggs were expressed into well sterilized watch glasses after which the milt 
was added. Before adding the milt a sufficient number of eggs were taken 
from the lot and placed in a fingerbowl of water, as a control. The 
fertilized lot was also placed in a fingerbowl and allowed to develop there. 
After the per cent. and character of impregnation was determined and the 
development well along in segmentation, changes of water sufficiently 
frequent to insure normal conditions for development were made. All 
dishes, pipettes, etc.. were thoroughly sterilized, first with hot water and 
then with 95 per cent. alcohol. Notwithstanding the fact that it was 
found that little danger of contamination existed, the precautions were 
strictly observed. In not a single instance was there any suspicion that 
the eggs were not fertilized by the desired sperm. 
1] wish to call attention to one defect in the methods of rearing the hybrid 
eggs. It may be objected that while the rearing of the eggs in the fingerbowl may 
be satisfactory for Fundulus and some other species it is not normal for a hybrid 
egg having a sperm from a species that has, for instance, pelagic mode of life during 
its deveiopmental stages. This unnatural condition may, therefore, in part at 
least, be responsible for the failures in the development, or even the particular 
stages at which development ceases. This objection, so far as we know, may or may 
not be of value. I see no way to avoid this experimental error, since it is not prac- 
ticable to cater to the demands of one of the parent species without, theoretically 
at least, infringing on the other. It may be said, however, that many of the species, 
especially those on which most stress has heen laid, have been successfully reared 
by this method, e.x., all the species of Fundulus, the two species of Sticklebacks, 
and the iwo species of Menidia. It is the belief of the writer that this objection 
may be disregarded. 
