-3- 
 
losses occurring within a certain time span coincident with a period of bad

weather are emergency losses, the resulting chances for error are very much

greater than would be expected from the kind of analyses of covey histories

that I have always made; in 1941-42, a winter of obviously depressed threshold

with evidence of severe predation losses occurring both before and after
the 
storm period, the losses during the storm period are ascribed solely to weather,

which simply isn't reasonable. Why should we not expect a substantial preda-

tion loss during the storm period, as well? My method of following covey

histories I would say is infinitely more precise than this, though I am 
hundreds of miles from the area and have only Gastrow' s notes with their

recognized limitations to go by. 
I agree wholly with your third paragraph on page 2, except for the im- 
plied emphasis on our present accounting for facts. It may not have been
clear, 
but in my last draft I tried to cut out explanations to the barest minimum;

I am willing to go still farther in this 0     , retaining only those explana-

tions that satisfy us both, unless, of course, you are willing for me to
sug- 
gest explanations strictly on my own stated responsibility. 
As to the fourth paragraph, rather than having separate sections on 
hypotheses and facts, I would favor leaving out all hypothesis. There still

seems to be some disagreement between us, however, as to what is hypothetical

and what is established as factual. 
I do not agree with you all the way in your fifth paragraph 4 page 
example of how hypothesis gets mixed up with facts. There seems to me little

room for question that the thresholds were low for 1936-37 (this is the least

certain case) and 1940-42. There is no doubt whatever that the gains for
the 
preceding su.;mers did fall below the "fideal'i curve, as they also
did in 1930, 
which did not precede a winter with a greatly lowered threshold. In the sum-

mary, I used ftconnection"1, which I think a poorly chosen word, but
otherwise I 
don't see that I assume a linkage that considering the actual evidence is
not 
justified by the word "association". But I emphatically do not
link "sub- 
standard gains and extra large winter losses" -- here again you lump
gross 
losses, emergency and nonemergency alike, in a way that is utterly fatal
to 
understanding of thresholds. 
The discussion of internal versus external properties that follows in 
the same paragraph is itself characterized by assumptions (especially concern-

ing the significance of movements of birds), which I don't feel are necessarI1

if at all valid. To go into this in any detail would be too wearisome and

probably unprofitable to attempt by letter; you are within your rights, never-

theless, to object to what you don't like, and I will be glad to consider
what- 
ever version would satisfy you. Before dropping this, I might suggest for
your 
consideration (not to be incorporated in the MS) that any factor that would

tend to increase irritability or intraspecific intolerance of higher vertebrates

on a regional scale could account for the "cyclic" depressions
of summer gains 
and winter thresholds in the populations with which I have worked; and the
pos- 
sibility of cosmic "master" influence still enjoys some favor in
modern thought 
(witness Elton's reference to astronomical processes on page 230 of his VVV

book). Nor,I may point out, would the operation of cosmic influence during

cyclic depressions be at all inconsistent with the concept of, and evidence
for, 
local traditions elsewhere advanced. 
I have a set of Albrechtts reprints and am also rather familiar with