44 



crystallise even, although in contact with unfiltered air, and 

 will remain clear, as was shown in the case of Sodic 

 Sulphate, for some seconds ; but on placing a minute dust of 

 the same substance on the hands, and rubbing them together 

 over the flask, crystallisation immediately ensues, as if by 

 magic. 



We may now vary the experiment by shewing that not 

 only is filtered air unable to effect crystallisation, but that 

 air which has passed through water or has been heated 

 strongly, is equally inert. 



Thus, by a suitable arrangement, w£ pass air down a 

 tube plugged with cotton wool, and then through a super- 

 saturated solution of Sodic Sulphate, which remains un- 

 affected, and air expired from the lungs, or, steamed air, 

 are alike ineffectual ; but if ordinary air be driven, by 

 means of a pair of bellows, into the liquid, crystallisation 

 follows. 



The foregoing experiments point clearly to dependence 

 on material particles in the air as a starting point of sudden 

 crystallisation, and it will be noted in the experiments above, 

 that whenever means were adopted to exclude or destroy 

 these particles, the air proved by itself impotent to cause 

 crystallisation. Moreover, it has been found that these 

 particles causing crystallisation must be themselves crystal- 

 line, and therefore the application of heat or steam to air 

 containing them, by destroying or dissolving the crystals, 

 renders the air inert. 



Again, a supersaturated solution of Alum is not affected 

 by UiC introduction of clean crystals into it — if these be of 

 different chemical composition — even though they be of 

 identical crystalline form. 



Thus, we wash crystals of different substances and drop 

 them gently into a supersaturated Alum solution, but no 

 result follows until, on introducing the minutest fragment of 

 Alum, crystallisation sets in from the nucleus and extends 

 rapidly through the solution. The resemblance between the 

 propagation of crystallisation and that of infectious disease 



