Sewage Purification. 255 
as soon as possible, surrounded preferably with ice till the 
moment that the process begins. A measured quantity is then 
taken and diluted with a measured quantity of sterilised distilled 
water, thoroughly mixed, and then a measured quantity of the 
diluted matter added to a tube of melted (of course sterilised) 
nutrient gelatine, then mixed thoroughly by a peculiar kind of 
oscillation, and the whole mass poured out on to a Petri’s plate, 
and instantly covered up. When the gelatine has set, the plate 
is placed in an incubator, and kept there at the constant tem- 
perature of 20° C. (68° F.) for two or three days, for ‘‘ colonies ”’ 
to form. These colonies appear as little beads in the trans- 
parent gelatine. By general agreement each colony counts as 
one organism, ¢.e. an organic unit without any reference to 
specific identity. The counting then begins. A black card 
ruled off into small squares of equal area is placed beneath the 
: plate of colonies; these squares are numbered. . One counts 
| generally with the help of a low magnifier the number of colonies 
in a square, and notes down the number on a correspondingly 
marked white paper. The total number in the several squares 
gives the number of organisms in the quantity of sample taken. 
In a crowded plate the enumeration is laborious, and in practice 
one doés not count all the squares, but takes an average of 
several, and multiplies that by the number of squares covered 
by the plate. I will now give you the results of some of my 
countings :— 
Experiment No. 6.— Date, April 8rd, 1894. Time, noon. 
Material, screened sewage. Average of two plates, 3,257,000 per 
cubic centimetre. 
Experiment No. 7.— Date, April 8rd, 1894. Time, noon. 
Material, effluent. Average number, 364,000 per cubic centi- 
metre. Note that the effluent contained something over 10 per 
cent. of the number of organisms that the sewage contained ; 
but as Mr. Figg tells me that he reckons the water takes two 
hours to get to the effluent, this number corresponds to the 
sewage of about 10 a.m. Possibly the 2 p.m. effluent would 
have given more organisms, as Mr. Figg says the sewage is 
generally thickest about noon. 
Experiment No. 8.—Date, April 4th, 1894. Time, 8.40 a.m. 
Material, sewage. Average of three plates, 1,269,000 per cubic 
centimetre. 
Experiment No. 10.—Date, April 9th, 1894. Time, 10.15 p.m. 
Material, sewage. Average of three plates, 8,297,000 per cubic 
centimetre. 
___ Experiment No. 12.—Date, April 13th, 1894. Time, 1 p.m. 
Material, sewage. Average of three plates, 11,000,000. I can- 
_ not explain this jump up. There had been no rain since the 
2nd. The temperature had, however, been abnormally high, 
a ta 
at 
a 
& 
¥ 
ts 
* 
4 
