PUTEEFACTION AND INFECTION. 105 



he based his first inference owed their barrenness, not to 

 the absence of Bacteria-geTuas from the air, but to the 

 inability or, rather, slowness of his mineral solution to 

 develop them. 



With regard to the part played by the visible motes, 

 I may repeat here what has been previously stated, 

 namely, that while the coarser particles could hardly 

 exist in their midst without loading themselves to some 

 extent with the minute germs of Bacteria, there is no 

 reason to think the motes indispensable for the diffusion 

 of the germs. Whether they are attached to each other 

 or not, the drjmess and the moisture of the air are shared 

 equally by both. The germs, moreover, float in the air 

 more readily than the larger particles ; and they, I doubt 

 not, when properly illuminated, shed forth a portion of 

 that changeless light to which reference has been already 

 made, and the perfect polarization of which declares the 

 smallness of the masses which scatter it. 



The prevalence of the germinal matter of Bactei-ia 

 in water has been demonstrated by the experiments of 

 Dr. Burdon Sanderson. But the germs in water, it 

 ought to be remembered, are in a very different condi- 

 tion, as regards readiness for development, from those in 

 air. In water they are already wetted, and ready, undei' 

 the proper conditions, to pass rapidly into the finished 

 organism. In air they are more or less desiccated, and 

 require a period of preparation more or less long to 

 bring them up to the starting-point of the water-germs.' 



' The process by which, an atmospheric germ is wetted would 

 be an interesting subject of investigation. A dry microscope 

 covering-glass may be caused to float on water for a year. A 

 sewing-needle may be similarly kept floating, though its specific 

 gravity is nearly eight times that of water. Were it not for some 

 specific relation between the matter of the germ and that of the 

 liquid into which it falls, wetting would be simply impossible. 

 Antecedent to all development there must be an interchange of 



