610 U. S. BUREAU OF FISHERIES 
erence to the date that approximately represents the mid-point of the time 
interval during which the samples were taken, this mid-point being the date 
of the median sample of the series taken during the interval covered. 
Disregarding for the moment the second and eighth frequency distributions, 
it should be noted that the sizes are grouped very uniformly around the mode 
at 37% and 38 centimeters in the first and third to seventh distributions, 
respectively. Let us call the mackerel around this mode “medium sized.” 
Turning now to the remaining two distributions (the second and eighth), we 
find the concentration of sizes represented by modes at 40% centimeters, in 
position with lesser concentrations at 45 and 46 centimeters. Let us call these 
the “large” and “ very large” mackerel. With these definitions in mind it is 
readily apparent from the graphs that medium-sized mackerel were caught in 
the southern fishery and during the summer New England fishery, while the 
large and very large-sized mackerel were caught during the cape shore season 
both along the coast of Nova Scotia and along the cost of New England by the 
few purse-seiners who remained in the New England fishery; and that, again, in 
autumn the large and very large mackerel appear in the catch off the New 
England coast. The full significance of these data on sizes can not be appre- 
ciated without a knowledge of the relative quantities of mackerel taken in the 
various portions of the season. We have not yet had time to analyze this 
feature in any accurate fashion, but we know that in general the quantities 
of fish caught during the southern fishery and the summer New England fishery 
were far greater than the quantities caught during the cape shore season and 
the autumn fishery, when the large and very large mackerel were present. With 
this in mind, it may readily be seen that the greatest bulk of the mackerel 
caught during the 1926 season were of medium size, and this may possibly be 
as high as 80 per cent of the total catch. We believe that it is highly significant 
that such a large portion of the catch was comprised of the medium-sized mack- 
erel with the mode uniformly at 37% and 388 centimeters. 
Referring now to the 1925 curves, shown by the broken lines,” there are cer- 
tain similarities and dissimilarities between the 1925 and 1926 data, which 
appear highly significant. From August 20 to October 28, 1925, correspond- 
ing to the latter part of the 1926 summer fishery, we have a similarly uni- 
fcrm group of sizes, the only difference being that the 1925 frequeney curves 
have their mode at 36 and 3614 centimeters, as compared with 38 centimeters 
in 1926. Furthermore, the frequency curves during the 1925 autumn season 
apparently correspond to the large mackerel of the 1926 autumn season, with 
their modes at 39144 instead of 4044, as in 1926. It probably is not venturing 
too much to assume that the shift in the modes of the medium-sized mackerel 
from 36 centimeters in 1925 to 38 centimeters in 1926, may be due to growth, 
just as the shift from 3914 to 40% centimeters in the large mackerel may also 
be due to growth. If so, it would appear that the stock of mackerel from 
which the bulk of our catch was taken in 1925 was of the same year class 
as that which formed the bulk of the catch in 1926. 
If this progression in size from one year to the next represents the growth 
of a Single year class from 386 to 88 centimeters, it is not unreasonable 
to suppose that in 1927 this year class would have grown so that its mode 
would be at 3914 centimeters and would then be of similar age and size 
to the large mackerel, with their mode at this length, in 1925. The 1925 
large mackerel have progressed from 391% to about 40% centimeters in 
1926. Considering these facts and assumptions, it is possible to trace, 
tentatively at least, the growth of the mackerel from 386 centimeters to 
88 centimeters to 3914 centimeters and to 4014 centimeters in four successive 
years. These four points, when plotted in the fashion of the usual growth 
curve, present a contour that suggests that they really represent the growth 
of the species. With the addition of the 1927 data, these assumptions prob- 
ably will be verified as facts. Meanwhile, it appears that our present catch 
of mackerel is based, for the most part, on a single year class, supplemented 
in small part by a year class of secondary importance two years older. 
This theory will be tested further by the examination of scales and otoliths 
from mackerel collected during 1925 and 1926, and these collections will be 
continued in future years. Age determinations of these scales and otoliths 
will be made entirely independéntly of the data on lengths, and should 
afford an entirely independent corroboration of the year groups tentatively 
discussed in this presentation. 

1 These have been plotted on base lines representing similar parts of the 1926 season. 
