188 THE FLOATING-MATTER OF THE AIE. 



§18. Change of Apparatus. Neiu Experiments 

 with Filtered Air. 



The source of possible error referred to in the last 

 section had been long present to my mind, and I had 

 already taken measures to avoid it. On the 2nd of 

 January, 1877, an infusion of turnip (sp. gr. 1006) and 

 an infusion of melon (sp. gr. 1008) were prepared and 

 introduced into a series of pipette-bulbs in the follow- 

 ing manner: — One end a, fig. 18, of a glass J-tube was 

 connected with an air-pump, the other end b was 

 closely plugged with cotton-wool, while to the third 

 branch of the J-tubs the neck of the pipette-bulb A 

 was attached by india-rubber tubing. A piece of the 

 same tubing, furnished with a pinchcock p, was also 

 attached to the free end of the J-tube beyond the 

 cotton-wool. 



The bulb A was exhausted three times in succession, 

 the pinchcock p being closed, and was three times filled 

 with filtered air, the pinchcock being opened. At the 

 third exhaustion the bulb was raised to a very high 

 temperature by a Bunsen flame, and finally filled with 

 filtered air. It was then plunged for a minute into ice- 

 cold water, from which it was afterwards removed, 

 detached from the J-tube, and then charged with the 

 infusion by means of a narrow pipette, / e, shown at the 

 top of B, fig. 18. 



The rationale of the above proceeding is this: — On 

 quitting the ice-cold water for the warmer air of the 

 laboratory, expansion of the air within the bulb would 

 occur. This would cause a gentle motion from within 

 outwards, opposing aU indraught of contaminated air. 

 The entry of the infusion into the bulb would, I thought, 

 also promote this outward motion. On the removal of 

 the pipette, which occupied but a very small portion of the 



