THERMAL RELATIONS. 



8i 



the germicidal action of the metal might arise. In no case should the cultures be 

 incubated in the tin tubes. When exposures are made in test-tubes of resistant 

 Jena glass, the cultures umst be lowered into the liquid air gradually, the fluid being 

 frozen from the bottom upward to avoid cracking the tubes. It requires about four 

 minutes to properly freeze a culture in a glass test-tube. Large volumes of culture 

 media should not be lowered into the liquid air, as it is wasteful, the air boiling away 

 rapidly. The writer began his experiments with block-tin tubes, as shown in fig. 

 67, but now uses tubes of Jena glass. The latter crack occasionally in spite of care. 



Fig. 67.* 



For very rapid freezing the amount of fluid in the tidje may be reduced to i cc. 

 Liquid air in Dewar glasses, and compressed oxygen, hydrogen, and carbon dioxide (?) 

 in steel cylinders may be had from the Eagle Oxygen Company, Incorporated, 121 

 West Eighty- ninth Street, New York City. The tanks of compressed gases may be 

 bought or rented. The following sizes may be had : Fifty gallons (280 pounds 

 pressure per square inch) ; 100 gallons (240 pounds pressure) ; 150 gallons (225 

 pounds pressure) ; and 200 gallons (280 pounds pressure). Cylinders may also be 

 had with the gas under much greater pressure. The cost of the oxygen is 2}4 cents 



*FiG. 67. — Dewar glass for liquid air, and block-tin test-tubes used in first low temperature ex- 

 periments with bacteria. About one-sixth actual size. 



