56 BACTERIA IN RELATION TO PLANT DISEASES. 
hydrogen or arseniureted hydrogen; it should therefore be passed not too rapidly 
through the following solutions in the order indicated: Saturated solution of lead 
acetate, 5 per cent solution of silver nitrate, 10 per cent potassium permanganate, 
10; per, cent: sodium -hydtate containing pyrogallol, distilled water. When ready 
for use the purity of the es die may be tested by burning i in test-tubes (mouth 
’ down , and also, if necessary, by the ordinary 
A : = raetlinils of. gas-analysis. ‘To avoid the evolu- 
tion of hydrogen sulphide the generator may be 
plunged into a jar of ice-water, as shown in 
plate 7. _Special care must be taken in sealing 
jars containing hydrogen, otherwise it will 
escape. In use, the gas is allowed to bubble 
slowly through the fluids into the culture- 
chamber, a large well-clamped Novy jar, the 
other tubular opening of which is connected 
‘air-tight with the tube of the vacuum pipe. 
The jar is first pumped out and the hydrogen 
is then allowed to enter. When the jar is full, 
the glass stopcock nearest it (at the left in plate 
7) is turned, and then, after allowing a few 
minutes for diffusion, the mixture of air and 
gas is pumped out. The vacuum cock is then 
turned. off and the hydrogen is again turned 
on slowly. This process is repeated five or 
six times, the gas being passed into the jar 
=. very slowly the last two times, so that it may 
oe be washed very clean. The Novy jar is then 
Fig. 52.*° '* sealed; disconnected, and set away in the dark. 
_ _ .The gas must,.of course, enter each wash-bottle through the long stem. It is 
desirable to have each wash-bottle two-thirds full of fluid, and there must be no leaks 
in'any part of the apparatus.} The hydrogen should be cut off before each exhaustion 
of the jar by turning the stop-cock nearest the jar.. The cock also should be turned 
off -before sealing glass tubes with flame and it must, of course, be known that the 
gas is free from admixture with air, otherwise an explosion will occur. — 
' It is easier to keep air out of gases than to remove it. The greatest care 
should’ therefore be taken to drive it out of a culture medium before it is inocu- 
lated. For the same reason gas should be allowed to flow for some time’ before 
it is collected-so'asto displace air which may have diffused into the generator and 
wash-bottles. This is also the reason why the water which is used to dilute the acid 
and the marble chips should be ~boiled.. If ‘there is much air mixed with the gas 
is not at all likely that a-single wash-bottle of sodium hydroxide and pyrogallol, 
*Fic, 52—Hempel’s simple pipette for liquid reagents used in gas-analysis. Breadth of stand, 
7 inches. 
+Consult a paper by Ewell, Centralb. f. Bakt., 2 Abt. III Bd., p. 188. 
