162 THE AMERICAN MONTHLY [July, 



more than ten species in any sample of water, especially when these 

 ai"e not species ordinarily met with, the water should not be used for 

 drinking purposes. He found that only in fifty-nine waters was this 

 the case, but that 169 waters contained more than 1,000 organisms per 

 cubic centimetre, sixty-six of these having over 10,000 (forty over 

 50,000) . From these figures it will be seen that some of the sources 

 of supply would be condemned by the old method but would be passed 

 by the new, and some condemned by the new would be passed by the 

 old. Migula found in all twenty-eight species, and in a series of tables 

 he brings out the fact that the number of colonies does not by any 

 means correspond with the number of species, though in some cases it 

 undoubtedly does so. This is, in fact, an exceedingly variable quan- 

 tity. It also comes out that putrefactive bacteria' are almost invariably 

 absent from spring water ; that they are most frequently found where 

 the number of species is great, and where the number of colonies is 

 between i ,000 and 10,000 per cubic centimetre; that they also occur 

 where the number of germs is below fifty per cubic centimetre, but 

 very seldom where the number is over 10,000. 



Dr L. Schmelk, who recently ( Centralbl. f. Bakt. und Para- 

 sitenk.^ Bd. IV^ No. 7, p. Jgs) pointed out there is a great increase 

 in the number of bacteria in the water supply of Christiana during the 

 period that the upland snows are melting most actively, now ( Cen- 

 tralbl. f. Bakt. und Parasitenk.., Bd. VIII., No. 4, p. 102) giAcs 

 further evidence collected during the last three years in proof of his 

 theory. The numbers he finds for those years were ten or fifteen per 

 cubic centimeter in March to 2,500 in April, 1888; 1,100 in 1S89, 

 and on March 28, 1890, 5,000; the breaking up of the winter snows 

 having occurred this year much earlier than usual. This is the period 

 during which the winter snows are melting, and after this is completed 

 there is no marked increase in the number of bacteria in the lake water 

 until the reappearance of the winter snows, some of the earlier falls 

 of which during October, November, and December melt and disap- 

 pear. In December the number of bacteria per cubic centimeter some- 

 times reaches 600, the highest point recorded during the year except in 

 March. Dr. Schmelk thinks that the increase is due to the action of 

 frost in breaking up the earth's surface, from which the contained or- 

 ganisms may be set free as soon as a thaw occurs and then washed 

 away along with the surface soil just as during great rain-storms. He 

 also points out that the masses of ice projecting into a river may form 

 "collecting" points for the particles suspended in the flowing water, 

 as more bacteria are always found in the water obtained from such ice 

 when melted than in the river water itself. He verified this by re- 

 peated experiments. He found, however, that when floating ice was 

 melting in water, though it contained a few more organisms than water 

 collected near the surface, it held far fewer than water taken from a 

 considerable depth. In the Christiana water-supply he found some 

 thirty species of bacteria, some of which occurred very seldom, some 

 at certain periods of the year only, and a few all the year round. The 

 amount of solids in the water varies from time to time, between 0.92 

 and 0.94 grammes per liter, and traces of ammonia can usually be 

 found in water during the time that it contains most bacteria. — -Supp. 

 British Medical Journal. 



