j 
SRA-FISHERIES LABORATORY. 131 
two local races, or of the distinction between summer and 
winter fish, is a matter for further research. 
We may say, then, that this method gives us good ground 
for supporting Heincke’s conclusions and opposing those of 
Matthews. But more research is needed on large samples 
from as many areas as possible. These samples must be fresh 
and in good condition; no method of preservation is 
satisfactory. Probably the characters on which most reliance 
can be placed are the number of keel-scales and number of 
vertebrae, but probably the main reason for this is that these 
characters can be determined more accurately than the others. 
In large samples of good fish it will probably be found that 
the other characters will give quite reliable results. 
I have tried to find out whether sex is of sufficient import- 
ance to make it necessary to treat the sexes separately. 
Unfortunately, it is difficult to get large series from the 
figures available, but the followimg results may be noted. 
Schley Herring : 
ee g 4 162 9: y? — 63475 2 n = 6 > P—0-28. 
Werrebrae; 92 3, 125 9: y7=6-775 -n—5 > P=0-15. 
Herring from Kiel, &c. : 
ee ota | (22-9: y2=6:201 : n—6 : P—0-28b. 
Wersebrae; 43 f, 52 o>. y?=1-012 : n=4 : P=0°8. 
Port Erin, 1914: 
mie 102g 0G) Oe wy? 4-679) n—5 > P=0-325. 
I have also examined the figures given by Duncker* for 
the sexes in plaice, which give the following results. 
Number of rays in dorsal fin : 
eng O18 oS v27— 15-81 n—17 : P=—0-46. 
Number of rays in anal fin: 
ites Olone. > ye— 15 -n—1le P—0-223. 
*Wiss. Meeresunters., N. F. IV, 1900. 
