of the Fislwry Board for Scotland. 
21 L 
SECTION B— BIOLOGICAL INVESTIGATIONS. 
I.-ON THE FOOD OF FISHES. 
By W. Ramsay Smith, M.B., CM., B.Sc. 
This Report gives a statement of a fourth year's investigations carried 
out on board the 'Garland' by Mr Thomas Scott, F.L.S. The results are 
set forth in a way similar to what was done in the case of the three 
previous years ; and, so far, they give a general indication of the relation 
of this year's investigations to those of previous years. But the pub- 
lished results give little idea of the amount of material available for study, 
both in Mr Scott's painstaking and. comprehensive returns, and in the 
abstracts I have had to make of his returns for the purpose of preparing 
the results published in the Annual Reports. The abstracts show not 
only the species on which the various kinds of fish feed, but also the 
numbers of fish examined at each particular station, the number containing 
food, and a list of the various species constituting such food. From this 
it will be evident that almost any question relative to the dietary of any 
species of fish at any station and for any month of the year, can be 
solved by consulting the abstracts. To make a comprehensive study and 
a digest of all those abstracts would entail an amount of labour, and such 
a lengthened report as is at the present time out of the question. I may say, 
however, especially since former Reports are being referred to and used 
for purposes of comparison by naturalists and others engaged in investi- 
gations elsewhere, that many things that are only hinted at, or given 
without much detail, can be elucidated by reference to the original 
abstracts. The value of a series of observations made with such 
accuracy as Mr Scott's have been, and made continuously at fixed 
stations for a period of four years, can scarcely be overestimated, 
especially when the numbers of fish examined are taken into account. 
Hitherto, when the subject of the food of any particular kind of fish has 
been referred to, the stomach of a single fish in many cases has furnished 
the sole evidence. More rarely, as in Mr Sim's investigations, a number, 
sometimes a fairly large and representative number, have been examined. 
Yet again, systematic investigation has been carried out at definite 
stations as in the case of Mr Cunningham's investigations at Plymouth, 
and Mr Holt's investigations on the Irish Coast. All these are useful in 
themselves as an indication of what fish in their native element may eat 
or do eat, also for comparison in various ways. But I would point out 
that the way of statistics is a very hard one, and one in which it is easy 
to stray. Putting aside for the moment the question of time and space 
distribution of fish food, and looking merely at the problem of ' What 
1 does a fish, say a cod, eat — to judge from an examination of a 
' number of specimens ? ' one has to face the question of the trustworthiness 
of the results and of the reasoning. I mean, Can one draw any trustworthy 
conclusion from ten fish? If not, then from a hundred? Or from a 
thousand? The law of statistics is very definite, although it is too 
often unknown or ignored. Let me apply it to a particular case, a 
