648 BULLETIN OF THE BUREAU OF FISHERIES. 
parisons are therefore necessarily limited to the whitefish production of those 
waters that lie within the borders of the Dominion of Canada and of the State 
of Michigan for the period 1892 to 1906, inclusive, and to the Great Lakes as a 
whole for the years subsequent to 1893. In the latter case, however, statistics 
are available only at five-year intervals. 
The distribution of whitefish fry is carried on at the expense of the state 
and national governments, and the annual official reports therefore contain full 
statements as to the number of fry distributed and the location of each plant. 
Whitefish fry have been distributed by the states named above with the exception 
of Illinois, Indiana, and Minnesota, and by both of the national governments, 
although the states of Ohio and Michigan have not distributed whitefish fry in 
recent years. 
By combining the statistics of the Michigan fisheries with those of the 
Dominion of Canada, complete returns are available of the catch of whitefish 
and of the nets used for Lake Huron for the years 1892 to 1906, inclusive. The 
exact number of pounds of whitefish brought in by each fisherman on this lake 
and the precise length of the nets used in the lake are recorded for each of these 
fifteen years. The plant of whitefish fry in this lake may be obtained by com- 
bining the figures of the Michigan and Canadian plants for the whole lake or for 
any part of it. We may thus study in this lake, or in any part of it, the effect 
of the distribution of whitefish fry on the catch of adult fish for a period of fifteen 
years. By comparing Canadian and Michigan waters for the same period, we 
may study the effect on the production of whitefish of certain restrictive enact- 
ments which have been enforced in Canada but not in Michigan. By similar 
methods we may study certain areas in Lakes Superior, Michigan, Erie, and 
Ontario, from which we have statistics from either Michigan or Canadian sources. 
In statistics of the catch of whitefish, figures of single years have little 
significance, since the catch of one year may be large owing to the weather con- 
ditions of that year having been favorable for fishing, while the catch of another 
year with which the first may be compared may be small owing to unfavorable 
weather conditions during that year, or the market price in one year may have 
been high, with a consequent stimulation of fishing operations, while in another 
year it may have been low, with a consequent curtailment of those operations. 
These annual fluctuations make it netessary to compare with one another not 
single years but periods of years. In the present paper market fluctuations 
are not considered, since the price of whitefish does not tend to fluctuate but 
rather increases steadily, while by comparing the average outputs of the three 
successive five-year periods comprised in the years 1892 to 1906, inclusive, an 
attempt is made to avoid the errors introduced by annual variations in the 
weather of the fishing season. 
