LARRABEE. — THE OPTIC CIIIASMA OF TELEOSTS. 219 



cient size to allow dissection with needles under the dissecting micro- 

 scope. They were killed and fixed in Kleinenberg's picro-sulphuric 

 solution, washed out in 70 per cent alcohol, and finally preserved in 85 

 per cent alcohol. The young cod were too minute to be dissected 

 in the same way as the trout. It was necessary, therefore, to imbed 

 these in paraffine, and section, using the compound microscope to follow 

 the course of the optic nerves. Vom Rath's platinic-osmic-acetiy fluid 

 was found to be the best fixative for use with the cod. 



Throughout this paper the writer will designate the optic nerves 

 according to the method employed by Dr. Parker, that nerve being 

 named right which runs to the right eye ; the one to the left eye, left. 

 For convenience, too, the specimens with the right nerve dorsal will be 

 termed rights, and similarly those with the left nerve dorsal, lefts. 



Results of Selected Matings in the Brook Trout 

 (fkdveUnus fontinalis Mitchill). 



First Generation. 



The first generation of the trout includes the young regarding which 

 the condition of the chiasmata of one or both parents, but not of the 

 grandparents, was known. The results obtained from this generation 

 are shown in Tables I, II, and III. In all, there are eleven lots, four 

 of which were obtained in 1901, the remaining ones in 1902. In each 

 year where a male or female was used for more than one lot, the fact 

 can be recognized by the correspondence of the numbers following the 

 sex signs. The numbers of the two years, however, relate to different 

 individuals. The control lot includes a number of young taken at 

 random from the fiy at the hatchery without any knowledge of the 

 condition of the crossing of the nerves in the parents. This lot, com- 

 prising young of various parents, is assumed to represent the average 

 condition of the trout in their natural state. 



The proportion of rights and lefts found in the eleven lots of the 

 first generation corresponds closely with that obtained from the control 

 lot. The percentage of rights in both cases is practically the same, — 

 56 per cent. With the exception of two lots, there is a slight excess 

 of rights in each case. In the nine lots where this excess is found, the 

 rights outnumber the lefts by from 0.4 per cent to 20 per cent. The 

 difference between the percentages of rights and of lefts in the series 

 as a whole is 11.8 per cent, which closely approximates the difference of 

 12.2 per cent found between the two kinds in the control lot. Nothing 

 in the net results indicates that in one case we are dealing with the 



