122 THE FLOATING-MATTER OT TUE AIE. 



tific work never lost faith in the substantial accuracy of 

 Pasteur. One striking example of his penetration has 

 an immediate bearing on the conclusion regarding 

 Bacterial clouds, independently drawn by me from the 

 deportment of the tray of one hundred tubes. On the 

 28th of May, 1860, Pasteur opened, on an uncovered 

 terrace a few metres above the ground, four flasks con- 

 taining the water of yeast. Nothing appeared in any 

 of them until the 5th of June, when a small tuft of 

 mycelium was observed in one of them. On the 6th a 

 second tufb appeared in another flask ; the two remaining 

 flasks continued intact and without organisms. On the 

 20th of July he opened, in his own laboratory, six flasks 

 containing water of yeast. Four of them remained per- 

 fectly intact, while two of them became promptly charged 

 with organisms. From these observations Pasteur in- 

 ferred the non-continuity of the cause to which so-called 

 spontaneous generation is due. This inference is quite 

 in accord with the notion of Bacterial clouds suggested 

 by my observations. Pasteur, in fact, sometimes opened 

 his flask in the midst of a Bacterial cloud and obtained 

 life, sometimes in the interspace between two clouds, 

 and obtained no life. 



Not with a view of repeating this observation, which 

 I had forgotten, but for another reason, I opened on 

 the 6th of January last a number of hermetically-sealed 

 tubes in one and the same room of the Royal Institu- 

 tion. The names of the infusions contained in the 

 tubes, the date of sealing them up, their condition be- 

 fore opening on the 6th, and their appearance six days 

 subsequently are given in the accompanying table. 

 1 chose for these observations tubes which contained a 

 little liquid in their drawn-out portions. In every case 

 the motion of this liquid, when the tube was broken, 

 indicated a violent inrush of air. 



