76 
importance in producing good year-classes. As there is a period 
of three years between the time of hatching and the joining of 
the commercial shoals in any great numbers doubtless other factors 
have their influence, but as they are thought to be of secondary 
importance they have not been considered yet to any extent. 
Figure II. shows graphically the number of barrels of herrings 
cured on the east coast of Scotland from 1862 to 1913, and the 
yearly values of the sum of the differences between spring and 
autumn maxima and summer and winter minima of tidal range 
as given in Table II. <A difference of four years between tidal 
data and herrings cured is taken. 
It will be observed that for a period of twelve years, from 
1867 to 1879 for herrings cured and from 1863 to 1875 for tidal 
data, the curves tend to run parallel. The greatest values for 
tidal data correspond with the greatest quantities of herrings 
cured, and the lowest points in the tidal curve are above similar 
points in the curve for herrings cured. For the following six 
years, 1879 to 1885, the curves are opposite in value, and the 
highest points for tidal data correspond with the lowest quantities 
of herrings cured and vice versa. From 1885 to 1891 (herrings 
cured) there is another period of six years in which the curves 
tend to be parallel, and this is followed by a six-year period in 
which the curves show practically the same as for the period 1879 
to 1885. The period 1897 to 1908 (herrings cured) can be taken 
as showing a general parallelism between herrings and tidal data, 
and that from 1908 to 1914 as having tidal data of highest value 
when the herring cure is at its lowest. 
From 1896 the actual herring catches for the east coast of 
Scotland are represented by the dotted line, and these emphasise 
the parallelism or convergency of the curves. 
In the two longer periods, 1867 to 1879 and 1897 to 1908, 
there is a tendency to change shown in the years 1873 and 1902 
respectively. There is also a doubt as to whether the year 1897 
or 1898 should be taken as the year of change. It is quite clear, 
however, that there is a periodicity in the relation between tidal 
activity and herring catches, and that the periods can be taken 
as approaching six and eighteen years in duration. Larger periods 
may exist and are suggested by the low values for tidal data 
coinciding with the comparatively small numbers of herrings 
