282 
BULLETIN OF THE BUREAU OF FISHERIES 
during April the percentage of fish showing a ring near the very edge of the scale 
increased steadily until all had it. At Halibut Cove the herring spawn during the 
last part of April and May. In a sample of spawning herring from San Diego Bay 
in December, the ring could barely be discerned on a small percentage of the fish. 
The scale reading presents some difficulties. As just stated, the ring seems to 
form about the time of spawning, which would indicate that it is due to the growth 
of the herring being almost or completely at a standstill at that time. The clearness 
of the ring would thus depend a great deal on how completely and for how long a 
time the growth was retarded. 
Thompson (1917) describes in some detail the difficulties that he encountered in 
attempting to read the scales of the herring in British Columbia. He mentions 
finding a winter zone in which the circuli seem to differ from the remainder. The 
inner margin of this zone tends to be marked by a “check” of some distinctness, but 
the winter mark is formed at the outer margin of this zone. He says that this is so 
often met with that it can not be regarded as otherwise than a normal phenomenon. 
Other difficulties were exceedingly wide summer zones, numerous more or less distinct 
“checks,” and the uncertainty of distinguishing the winter rings near the margin of 
scales of older fish. 
We have met with similar difficulties, but judging from Thompson’s text and 
plates the scales of Alaska herring are very much clearer than those of British 
Columbia. The scales of the San Diego Bay herring are slightly less clear than 
those of Alaska. 
One difficulty met with in the very old fish, besides the difficulty of distinguish- 
ing between the marginal rings, is that of distinguishing the first and often the second 
annual ring of the scale. Whether this is due to the more numerous layers of scale 
substance in the older fish changing the refraction of light is uncertain. 
All of the samples contained fish with well-defined rings on their scales, likewise 
all of the samples showed fish with regenerated scales, which are worthless for age 
determinations. The proportion of regenerated scales on the individual fish varied 
greatly, so that in reading the age of an individual from two to a dozen scales were 
usually examined, this often being necessary in order to obtain ones that were fully 
legible. For this preliminary report we have read chiefly scales from central Alaska. 
(Table 33.) 
GROWTH RATE 
A knowledge of the rates of growth in the various localities is desirable because 
of its many applications. Whether the difference in average size between any two 
localities is due to a difference in rate of growth or to a difference in the age composi- 
tion (poss : bly due to a reduction of the numbers of older fish in one of the localities) 
is important in studying depletion. The differences in growth rate are also a valuable 
aid in studying the independence of areas, as, where the differences are great, it 
shows a lack of migration. For instance, the differences in growth rate between 
Unalaska and Stephens Passage herring are enormous, those at Unalaska (Dutch 
Harbor) at 6 years of age being 6)2 centimeters greater in body length than those of 
Stephens Passage. 
To grasp the real significance of these differences in growth rate the comparisons 
should be made on the basis of weight, as any given increase in length represents a 
far greater amount of growth in a larger fish than in a smaller fish. The weight 
varies a great deal within the same locality at different seasons but only slightly 
between different localities at comparable seasons of the year. Therefore, not 
