190 THE FLOATINQ-MATTEE OP THE AIE. 



maintained for ten minutes at the temperature of boil- 

 ing water. After a lapse of twelve hours their sealed 

 ends were broken off by means of a file. 



In our experiments on the 28th of December turnip 

 and melon subjected to ten minutes' boiling entirely 

 gave way. In these present experiments, where greater 

 care was taken, two out of the six turnip-bulbs and 

 three out of the six melon-bulbs remained permanently 

 barren. Even this amount of success proved afterwards 

 so exceptional that it might be fairly regarded as 

 accidental. 



On the 4th of January the experiments were con- 

 tinued. The pipette-bulbs employed were first carefully 

 washed with carbolic acid, which was removed as far as 

 possible with ordinary water. They were then washed 

 with a solution of caustic potash, and finally rinsed out 

 with distilled water. They were not subjected to the 

 action of the Bunsen flame. The infusions employed 

 were turnip (sp. gr. 1006) and melon (sp. gr. 1008), 

 eight bulbs being filled with each infusion. 



I could not be certain that the motion of the liquid 

 fillet at the end e of the pipette with which the bulb B, 

 fig. 18, had been charged, had not drawn into the neck 

 of the bulb a modicum of the external air. In the pre- 

 sent experiments, therefore, the method of charging the 

 bulbs was modified in the following way: — The glass 

 J-^ube employed in our last experiments had its end a, 

 which was to be connected with the air-pump, drawn 

 out to a small orifice and bent as in fig. 1 9. The branch 

 connected with the bulb was also drawn out to a tube 

 of fine bore, which entered the neck of the bulb for 

 some distance to c, the thicker part above being con- 

 nected with the neck of the bulb by india-rubber 

 tubing. The end b, as before, was plugged with cotton- 

 wool and provided with a pinchcock, p. The object 



