130 THE AMERICAN MONTHLY [July, 
This fact impressed me so strongly after a number of observations that it be- 
came possible to anticipate quite accurately the number of bacteria present 
by looking at the water in the tube with transmitted light. 
It seemed reasonable to conclude that whatever agency brought the sus- 
pended earth also brought the bacteria, and that the earth contained the bac- 
teria. Throughout the winter of 1886 I noticed that after heavy rains the 
turbidity increased quite suddenly, this fluctuation, of course, producing a 
corresponding rise and fall in the number of bacieaet The an washing 
down the soil from the surface drained by the tributaries of the river, was 
thus the cause of the turbidity. But was there any relation between rainfall 
and the number of bacteria? Through the kindness of the Signal Office I 
obtained the data given in the third column of the table. Comparing the 
second and third columns, the relation is certainly not on the surface. The 
heaviest rains occurred in July, but the number of bacteria did not rise per- 
ceptibly, and no turbidity appears. The only explanation which suggests 
itself is that which must be sought in the changed condition of the surface of 
the soil in winter and summer as regards vegetation. The precipitated water 
is caught by the foliage of trees, by the grass and herbage, which clothes the 
soil everywhere. The soil itself is at the same time more firmly bound to- 
gether by the vegetation itself. In winter all this is changed. The absence 
of vegetation leaves the loose soil ready to be washed into streams by rain 
and melting snow, carrying with “it the bacterial vegetation. 
The majority of bacteria carried into the river are, no doubt, harmless, but 
what is to prevent the infectious micro-organisms of typhoid ‘and other dis- 
eases from being washed down and carried into our houses with the sus- 
pended matter? The danger is thus not constant, but only occasional. The 
number of bacteria may have no direct significance, but it is certainly an 
index of the possible danger. It is safe to assume that Potomac water, free 
from suspended matters, contains from 50 to 200 bacteria in tcc. This will, 
no doubt, be found a low average for unfiltered river water when more statis- 
tics have been collected of other streams w hose water is used tc Beli ce towns 
and larger cities. 
No qualitativ e examination of the different kinds of bacteria was ‘ae for 
want of time. Liquefying bacteria were constantly present ; when the bacteria 
were few in number, as in summer, as many as 50 per cent were liquefying, 
so that counting was somewhat difficult, and many plates were lost by the con- 
fluence of the large colonies and the total liquefaction of the gelatin layer. 
When the number was very high, as in winter, the liquefying forms did not 
increase in the same proportion, but formed only 5 to 10 per cent. of the 
whole. These observations led to the inference that they are constant inhabi- 
tants of the water, and that attention must be directed to them, first of all, 
if individual forms are to be more closely examined. 
O 
Chinese fish lines.—A communication from Miss Adele M. Fielde states 
that near Swatow, China, the silk-glands are taken from the larve of several 
species of large lepidopterous insects just before they enter the pupa stage and 
are made into fishing lines. At this period in the life-history of the insect the 
glands are full of the viscid white substance from which the cocoon is to be 
spun. The silk-glands of a species of A¢/as were found to be one yard long, 
a tenth of an inch in diameter at the free posterior end and a hundredth of an 
inch in diameter at its anterior end. The two glands extend nearly the whole 
length of the body cavity on either side of the alimentary canal, lying in loops 
of varying length and uniting in a single duct under the mouth, as in the silk- 
worm, Bombyx mort, The Chinese make a transverse cut across the back of 
