76 BACTERIA IN RELATION TO PLANT DISEASES. 
The writer has no very satisfactory way of making exposures for determining 
the minimum temperature for growth. His method is to make such exposures in 
the bottom ofa large, well-filled ice-box, which is opened as little as possible during 
the progress of the tests, and then only for the briefest periods. The degree of cold 
is governed by the amount of ice. A good thermometer is exposed in the midst of 
a bundle of inoculated, tubes, and if the temperature shows any tendency to rise 
more ice isadded. Under the most favorable circumstances the temperature of the 
*Fic. 63.—Modification of the Ostwald water-bath used by the writer for thermal death-point 
experiments. This consists of a porcelain-lined pot 11 inches in diameter at the top. ‘This is filled 
with water kept in motion by a water-wheel turned by electricity. The heat is applied by means of a 
Friedburg burner and is controlled by Roux’s thermo-regulator. Murrill’s gas-pressure regulator 
is shown at the left. 
