17 



have been made, and there is no exception to the facts plainly 



exhibited in the table. The actual size is not very different from 



the calculated size. In some cases they are exactly in agreement. 



But it is plain that in every case the actual size at the first two 



winter rings is larger than the calculated size, at the third wdnter 



the sizes may be equal or somewhat greater or smaller, and at 



the fourth winter ring the calculated size is nearly always greater 



than the actual size. As will be seen there is no exception to the 



results thus stated in general terms. They are as constant as the 



apparent contraction in size revealed by calculation. There is 



no doubt either that the two results are causally related. 



TABLE IV. — The Average Actual and the Calculated Sizes 

 OF Herrings, 



It is evident therefore that the scale does 

 at the same rate as the fish. The scale is not 

 herring has attained a size of about 4 cm. It 



not grow exactly 

 formed until the 

 appears late and 



