486 



BULLETIN OF THE BUREAU OF FISHERIES 



in the species. A study of the rate of growth of specimens from both situations 

 would, of course, also aid in determining their racial identity. In Table 69 are 

 compared extensively 10 specimens from Green Bay — 5 from the shore and 5 from 

 the deepest water — and 9 taken at various ports on the lake. 



Size variations. — Below are given, for the characters that vary most with growth, 

 values for large fish (225 millimeters and over) and small ones (under 225 millimeters). 

 More detailed figures are shown for some of these characters in Tables 8 to 11. 

 Ten specimens under 200 millimeters are compared extensively with larger fish in 

 Table 69, also. 



L/H: 



Large fish, (4.1) 4.3-4.5 (5). 

 Small fish, (4) 4.2-4.5 (4.6). 

 H/E: 



Large fish, (3.6) 4-4.2 (4.7). 

 SmaU fish, (3.5) 3.7-4 (4.3). 

 H/M: 



Large fish, (2.5) 2.7-3 (3.3). 

 Small fish, (2.4) 2.7-3 (3.1). 



H/S: 



Large fish, (3.3) 3.7-4 (4.4). 

 Small fish, (3.4) 3.6-3.9 (4.1). 



Pv/P: 



Large fish, (1.6) 1.9-2.2 (2.6). 

 Small fish, (1.6) 1.8-2.1 (2.5). 



Av/V: 



Large fish, (1.4) 1.6-1.8 (2.3). 

 Small fish, (1.3) 1.5-1.7 (2). 

 L/D: 



Large fish, (3.6) 4-4.9 (5.3). 

 SmaU fish, (4.1) 4.4-5 (5.8). 



Jaw: 



Large fish, equal or shorter, 92 per cent. 

 Small fish, equal or shorter, 65 per cent. 



It appears from these data that the eye and paired fins decrease proportionally 

 in size with growth, the depth of the body becomes greater, and the jaw tends to 

 become shorter. The head and snout do not change markedly in relative size, but 

 they also appear to be a trifle shorter, relatively, in adults. 



A few individuals of both sexes have been found sexually mature at 165 milli- 

 meters, but usually those smaller than 180 milhmeters show no development of the 

 sex glands. 



COMPARISONS 



Artedi is usually easily distinguishable from any of the deep-water Leucichthys. 

 Its flesh is much less fat that that of any of these, and it is usually distinguishable 

 from them by other characters also. Small individuals, however, sometimes closely 

 resemble hoyi. A discussion of the differences between artedi and hoyi is found on 

 page 452. The differences between artedi and johannx, alpense, zenitMcus, reighardi, 

 nigripinnis, and Jciyi, are given on pages 353, 366, 391, 403, 419, and 436. 



GEOGRAPHICAL DISTRIBUTION 



The herring of Lake Michigan occur in schools and are taken extensively at some 

 seasons of the year out of virtually every port on the lake. The long stretches of 

 sandy shore seem favorable for this species, and herring are abundant enough therefore 

 in most places to make their capture profitable. Green Bay is the most productive 

 area. I have collected specimens from virtually every port visited. The data for these 

 are given in Table 68 and are platted on the lake chart in Figure 4. 



METHODS OF CAPTURE 



Most of the herring are taken in gill nets and pound nets with small-meshed pots. 

 Gill nets are in use in Green Bay in the summer when the fish are offshore, in the 



