476 SEMI-CENTENNIAL OF Токвкү BOTANICAL CLUB 
ferent flowers as 35.4 X 34.2, 39.3 X 36, and 36.6 X 33 microns; 
the similar grains from the median (i. e., shortest) mid-length 
stamen of the short-styled flower showed such averages as 34.8 X 
32.7, and 38.1 X 33.6 microns. The average for a single flower of 
grains from the longest stamen in the mid-styled form ranged 
from 42.3 X 37.2 to 49.5 X 42.3 microns; while the small grains 
from the shortest stamen of a single long-styled flower furnished 
such averages as 22.8 X 23.1, 24 X 21, and 25.8 X 23.1 microns. 
The proportions of the grains are by no means constant, and in 
examining many grains one gains the impression that the two (or 
really three) diameters balance each other even more than is 
indicated in these sample averages, that is, that when one diameter 
increases beyond the average, the other decreases correspondingly 
so as to keep the volume average more constant than appears. 
It should be remarked that the above measurements were all 
based on material preserved for a year in formalin, and that they 
uniformly average less than a small number of measurements made 
in August 1916, of grains taken from fresh flowers and mounted in 
water. These few measurements of fresh grains, considered not 
enough to be reliable, in general approach more nearly the dimen- 
sions of a series made by Halsted* whose report came to my notice 
only after my own had been completed. As indicated by 
Halsted, dry pollen from fresh anthers is so contracted as to 
make its measurement of little significance. It might be expected, 
however, that though the amount of swelling of the grains would 
be greater when mounted in water than when fixed in formalin, 
nevertheless it would be proportional in similar grains, whatever 
the medium used; and as a matter of fact the ratio between my 
large- and mid-size grains is almost exactly the same in fresh and 
preserved material; a greater difference between the size of large 
and smallest grains mounted in water as compared with similar 
grains in formalin, I assume to be chargeable to the small num- 
ber of measurements of fresh grains. 
In order to obtain a comparison between our ellipsoid pollen- 
grains and those reported by Darwin as spherical, the measure- - 
ж Halsted, B. D.  Pickerel weed pollen. Bot. Gaz. I4: 255-57. 1889. Inthis 
brief article, no indication of the number of measurements is to be found; . with the 
mere statement that ‘‘only three prevailing dimensions" occur, and E VE of 
faulty calculations, it seems worth while to detail my own — 
