80 J. H. ORTON. 
counted separately from the fin-rays posterior to them ; thus the records 
take the form of a+b, the sum of which gives the total number of rays 
in the fin. A cursory examination of the records indicates that 24-17 is 
the commonest form for the dorsal fin and 2+-15 the commonest form for 
the anal fin (Figs. 4 and 5). 
In the case of Character 17, which is stated in the scheme to be the 
NOT % 
Fie, 2.—Diagram of Tail-bones of Herring, from Williamson. A is the Jast vertebra 
counted in the present work. B is regarded by Williamson as the last vertebra. 
Not != Notochord ? 
“serial number of first vertebra having complete hzmal arch,” the 
records have been made in the form of (a+b) where b=the total number 
of vertebrae with complete hemal arch—not, however, counting the 
Fic. 3.—Vertebre, showing incomplete but ‘‘ well-developed” hemal arch, 
terminal vertebra-like ossicle regarded by Williamson as the “ last 
vertebra” (see Fisheries, Scotland, Sci. Invest., 1914, I (April, 1914) 
Fig. 7, B, p. 21). Williamson’s figure is here reproduced as Fig. 2. 
In this character (17), a=the number of vertebree having an incomplete, 
but “ well-developed ” arch, and an arch was considered “ well developed ” 
if the heemal processes were almost as large as those of the first complete 
arch, and if these processes possessed even the smallest trace of an internal 
cross-piece (see Figure 3). It should be mentioned that all inter- 
