80 BACTERIA IN RELATION TO PLANT DISEASES. 
tious material used should be the thinnest obtainable film of fluid across a carefully- 
measured 1 mm. loop, so as to avoid crowding the plates. ‘The same loop should 
be used in all cases, and it should be dipped into and out of the fluid always in 
the same way. After pouring, set the plates on a perfectly level spot (fig. 66), until 
the agar has hardened. If the work has been well done, there should develop 
an approximately uniform number of colonies in each plate. ‘The tubes of inocu- 
lated water or bouillon are then immediately lowered into the liquid air and exposed 
to it for the predetermined time, after which six additional Petri-dish poured plates, 
of the same size and inoculated in the same way, are made from each tube for 
comparison with those prepared before the exposure. ‘The tubes may be thawed out 
by exposure to the air for three minutes and then to tap-water for five to seven 
minutes. The exposures are best made in Dewar glasses (fig. 67). When the 
exposures are long, a loose tuft of absorbent cotton should be placed in the mouth 
of the glass, or it should be covered with a hair-cloth cap, to prevent excessive 
- Fig. 66.* 
evaporation. Under these conditions the air remains liquid for a number of days. 
At first the temperature is about minus 190° C., rising gradually to minus 180° C., 
since the nitrogen evaporates somewhat faster than the oxygen. The glasses are 
fragile and should be handled carefully, especially when filled with the air. As long 
as they contain liquid air it is safer to keep them in their containing-case, packed 
about with cotton or felt. One should be careful to avoid cracking the inner wall 
of the glass, as might happen by dropping some hard substance into the receptacle, 
otherwise an explosion will occur, the space between the two walls of the Dewar 
glass being a very perfect vacuum. 
When the exposures are made in block-tin tubes, the culture should be frozen 
at once on pouring into the tube and the second set of plates should be made 
as soon as the fluid has thawed, z ¢., within about ten minutes, for which purpose 
the culture should be poured out into a glass tube, otherwise complications due to 
*Fic. 66.—Leveling (nivelling) apparatus for use in making poured plates. About one-third 
actual size. 
