326 BULLETIN OF THE BUSSEY INSTITUTION. 
eaten, as stones; either after the pulp had been removed by fruit- 
eating birds or when the pulp had decayed or dried up after the 
cherries had fallen from the trees to the ground. The correctness 
of this supposition was fully verified by the facts that wild-cherry 
stones were found on the ground as late as November 5, and that 
through all the autumn and winter months the pigeons continued 
to void upon the roof dung that contained a large proportion of 
broken cherry-stones, at times, that is to say, long after the fruit 
which formerly enclosed the stones had wholly disappeared. 
At frequent intervals during six or seven months I carefully in- 
spected the roof and continually collected fresh portions of the 
dung charged with fragments of cherry-stones from a place acces- 
sible to no one but myself, which was invariably swept clean after 
each collection had been made. The persistence of the appearance 
of the cherry-stone chips throughout the entire winter, even in 
times of snow and of severe frost, was so very remarkable that I 
am induced to print an abstract of my notes taken at the moments 
of observation. It will be seen that numerous fragments of cherry- 
stones were contained in the dung at all times and that it was only 
when the ground was pretty thoroughly covered with snow that any 
appreciable diminution in the number of the chips was noticeable. 
It is still a matter of surprise to me that the birds could have any- 
where found so many cherry-stones during so long a period, par- 
ticularly as they must have been subjected to the competition of 
squirrels and field-mice, which are by no means rare in the vicinity. 
It should be premised that after the middle of October the birds 
were less frequently to be seen on the building than had been the 
case during the two previous months. As the weather became 
cool they changed their resting-place a few feet to a sunny nook 
where they were to be seen occasionally during the remainder of 
October and throughout November, December, and January. 
During February and March the visits were more rare, and they 
wholly ceased after April. 
Collections made on the 5th, 10th, 15th, 20th, agd 25th of No- 
vember were abundantly charged with the broken cherry-stones, 
and the same remark is true of collections made at intervals in 
December. ‘The weather remained open until January 1, 1882, on 
which day the roof was cleared of dung that contained many frag- 
ments of the cherry-stones. That night six or eight inches of snow 
fell which was badly drifted. The ground was, however, covered 
for the most part, and although there were some places which the 
wind had blown bare they were not frequent. Very cold weather 
