CONFIETIAL: NOT FOR PUJBLICTION. 
 
 
Copy of memorandum given to ,Ur.Hoyes Lloyd, August 1931, by Charles 
Elton. 
 
These statements are based mainly on Hudson's L3ay Company enquiries 
and will be published in 1932. Pa-ts are already available in .natamek 
Conference RePort, in press. 
1. The H.B.Co. muskrat sales for all Canada show a marked 10-year cycle 
    The peak years of abuntance (biological, not sales) were 1871-2, 
                                                             1881-5 
                                                             1091-2 
                                                             1902-3 
                                                      probably 1912 
                                                             1921-22 
    Smaller peaks In 1695 and 1903. Vain cycle is clear and inverse to 
    snowshoe rabbit cycle whose peak years in central Canada were 6-05, 
    14-155, 24-25. 
 
 2. In the last cycle tne fur post data were analyzed (1916-1927). TLhere

    was a definite central area with regular cycle. Outsidle this there 
    was irregularity or disagreement. 
 
 3- For 1916-1931 tuere was an inverse cycle in'tae southern prairie 
    region of .iiddle 'West, best .roved for Saskatchewan. (This available

    in Bradshawe's game reports, Dept.AGric. \ebina) 
 
 4. There was thus a switch, with border roughly in Grove ielt. 
 
 5. Virtually certain that muskrat cycle connected with water-levels in 
    sloughs etc. Low water kills them by droa,1ht and freezing. 1ut this

    general hypoothesis not yet confirmed by water-level data (not yet 
    properly worked up), except in particular instances of muskrat 
    destruction. The water cycle not yet properliy demonstrated on a 
    long series of data. 
 
 6. This switch coincides exactly with tne present duc.-droinat map, and

    the distribution of water-levels accords with tne 1921 iuskrat era. 
    There were southern droub-hts then, followed by water and muskrats 
    and again drought and few muskrats in 1931. 
 
 7. If we can argue from muskrat cycles to water level cycles (not yet 
    proved properly, out very reasonable hypothesis) certain predictions

    can be Lade. 
 
(. After 1931-32(-33) the northern muskrats will start to decrease. 
      This may imply correspandinG oad conditions for duck In tae north-

      ern forest region. 
(b)   During th same period t-e sout ern area snould improve again 
      until about 1938. 
(c)   In other words the breeding status of those ducks may in ta±e
past 
      aave been made posaL ble by tae fact that there was always p    pf

      their breeding area in ood shape. 
(d)   But we must allow for two other factors: 
 
     (1) The actual positions of the areas may shigt as probably suown