3o8 BULLETIN OF THE BUREAU OF FISHERIES. 



the time of formation of the last amiulus. In the case of the comparisons of measured 

 and calculated lengths (table 2), if the annuli represented winters a discrepancy between 

 the averages would be expected; for the fishes measured would be approximately an 

 even number of years old (spawned in June and measured in July and August), while 

 the winters represent points midway between birthdays. But the calculated and 

 measured lengths agree remarkably (table 2; also fig. 2 and 7), suggesting the view that 

 annuli are year-old marks, but not winter marks. 



These observations seem to justify the opinion that the annuli are simply the 

 margins of the laminae composing the scale. Plate li, figure 3, a photomicrograph of 

 a part of a scale under high magnification, including two annuli in the field, shows 

 clearly that the annuli are not circuli closer together, but simply branching circuli. 

 This branching possibly may be explained by the disproportionate growth of the an- 

 terior and posterior fields. 



It is seen that while several circuli are being formed across the anterior field, only one 

 is formed on the lateral field, hence the branching. A glance at plate lvi, figure 19, will 

 show that all the circuli on the anterior field are traced back to a point on the posterior field 

 where they join the following circulus, and that an annulus is being formed continually. 

 Thus the last circulus on the periphery is, at its posterior extremity, part of an annulus 

 that will not be complete till the next year. The beginning of a new annulus appears 

 to be determined by a sufficient lateral growth to permit the formation of another cir- 

 culus. 



But this does not explain the annulus on the anterior field. Here, as elsewhere, it 

 seems to be the edge of a lamina. Just why these laminas end at the ends of years is 

 yet undetermined. It suggests that the fish passes through year cycles of growth, and 

 that one lamina is formed each year. It has been suggested that the fish spawns every 

 year from the first, and that the laminae represent differences of calcification during 

 spawning time. 



Dahl proposes a unique theory. The scale is secreted by the floor of the scale 

 pocket and the increasing size of the pocket explains the increasing size of the scale. The 

 thickness of the scale is only dependent on continued secretion. Thus if the scale 

 pocket remains constant in size, so will the scale; if it increases in size, a new ring will 

 be added to the periphery, etc. But a contraction of the pocket will produce an upward 

 fold in the thin edge. It is conceivable to him that during the spawning period the 

 body is more or less distended and while it is in this condition a new layer is added from 

 the floor of the scale pocket. Now a contraction of the body follows the spawning 

 period and with it a contraction of the scale pocket, pulling the thin periphery upward. 

 Such a process would produce structures similar to what we know the annuli to be, if 

 we leave the interior structure out of consideration. 



The evidence, however, does not point to this conclusion. That spawning has noth- 

 ing to do with the formation of annuli is evidenced by four different points: 



(i) Annuli are often found running in a direction contrariwise to that of the 

 circuli. 



(2) Spawning leaves a mark on scales of a character quite different from that of 

 the annuli. (Masterman 1913a, Milne 191 3, Calderwood 19 11.) 



(3) Expansions and contractions of the body of the fish consequent upon the spawn- 

 ing of the fish could not possibly affect the scales on the caudal peduncle, head, etc., yet 



