April, ’24] 
ENTOMOLOGICAL STATISTICS, DISCUSSION 
209 
conclusions, these details will be available to substantiate your work. 
Do not insist on their being published. I am one of the editors of a 
research magazine—that of the Department of Agriculture—and I 
know what comes over my desk and the burden of detail often gives me 
great pain. 
In conclusion, and referring again to the widespread doubt and sus¬ 
picion which attaches to statistics and statisticians, I will venture 
another old story, namely, the definition of liars: “Liars, damn liars, 
and statisticians.” 
Mr. H. A. Gossard: I noted in my memorandum, as the speakers 
were going along, some of the things Dr. Marlatt brought out, and I 
am glad he brought those things out so distinctly. I will not now need to 
say very much about them, but he is the last speaker on the floor, and, of 
course, he is leaving the last impression. I hardly agree with him to the 
full degree of his seeming condemnation or perhaps I would better say 
the limitations he is disposed to endorse; but I heartily agree with him 
about the personal element that so vitally affects all conclusions and 
forecasts. I recognize in this method of taking statistics one of the 
methods by which we may hope to make progress. It is worth trying. 
Don’t trust it too far, but it is one of the most promising programs we 
have for making real progress. 
I thoroughly agree with Dr. Marlatt that there is very much in the 
personal element, and while I won’t say there is such a thing as an in¬ 
stinctive Entomologist, who knows and acts by instinct, there is some¬ 
times something that is quite akin to it. Yet, I usually find that the 
man who is operating apparently by instinct, really has in the back of his 
head somewhere, some statistical data, or some experience which is the 
basis for his judgment. Now, for instance, when we come to the matter 
of entomological forecasting—we have done considerable forecasting at 
one time or another in our state in regard to the coming of Hessian 
flies; we add something to the method that Mr. Larrimer has pre¬ 
sented—in fact, we have all of his data placed at our disposal. But we 
have added to that for the past several years, a wheat survey made just 
before the harvest time, by which we take samples from several fields in 
each county and make an accurate count of the degree of infestation. 
But we don’t wholly trust that count. We don’t pretend that the 
figures we get are absolutely accurate or representative, but we know 
they are approximately accurate, within 10 or 15 per cent, anyway, if the 
infestation is high, and much closer if the infestation is low. We know 
