212 BOTANICAL GAZETTE | SEPTEMBER 
size of a walnut, and fell with sufficient force to break windows. The 
stones were carefully washed, melted, and plates made from the water. 
The number found was 628 to 729 bacteria per 1°. Only bacteria 
were discovered, and neither fungi nor yeasts were noticed. He 
described the morphology and cultural characteristics of two cocci, 
and two bacilli; of these one coccus was pathogenic in large doses to 
rats. 
On July 15, 1897, a violent hailstorm swept over Guelph, and hail 
fell heavily for about fifteen minutes, with such force as to break many 
panes of glass in the greenhouses. The stones averaged 20” in 
diameter, and looked like round bullets made in a badly fitting 
mould, with a projecting rim around the center. A number of these 
were taken to the laboratory, washed in mercuric chloride (1 to 500), 
rinsed several times in sterilized cold water, and each stone thus 
treated was dropped into a tube of melted nutrient gelatine, thoroughly 
shaken, and plates then poured in the usual manner. Four days after 
the plates were counted by the aid of Pakes’ apparatus; the quantita- 
tive results were as follows : 
12 per hailstone, all bacteria. 
ae ae “ec oe 
35 
eae a a few molds present. 
Kok Dal * nearly all molds. 
368 * . a few molds 
608 “ - mostly molds 
9s) tad ty a few molds. 
1280 “ - “ « 
Fa4ag “ «6 “ 
2880 “ “ “ 
3680 “ “ « “ 
11464: an average of 955 per hailstone. 
These numbers err on the small side, as a portion of the hailstone 
was lost in the cleansing process. All the bacteria and a number of 
the moulds were isolated, and their cultural characteristics noted. 
Among those present were Penicil/um glaucum, Mucor sp., Aspergillus — 
sp., B. fluorescens ligquefaciens, B. fluorescens non-liquefaciens, a protean 
form similar to Proteus vulgaris (Hauser), and one other germ subse- 
quently described. No micrococci were found. 
On July 30 another hailstorm occurred more violent than the 
