STUniKS ON FRiaiENTATION, 87 



point of the \\'iro we withdraw the glass stopper which closes 

 the india-rubber tube on the right-hand neck of the flask 

 (Fig. 14) and drop the wire through that tube ; we then I'eplace 

 the glass stopper, after having, by way of additional precaution, 

 passed it rapidly through the flame of the lamp. There is no 

 doubt that we expose ourselves to error in consequence of 

 having to convey the wire through the surrounding air, and 

 also, in consequence of having previously to open the flask ; 

 but, as we have already remarked, this double cause of error has 

 never, we may say, interfered with the exactness of our experi- 

 ments, the volume of air. with which we are concerned being 

 exceedingly limited. Moreover, our flask being in free com- 

 munication with the exterior air, by means of the opening in 

 the curved, slender tube, there is no inrush of air when we 

 withdraw the stopper. The chance of encountering a spore or 

 fecund germ and introducing it into the flask on the wire that 

 is charged with the others, is so remote that we have considered 

 it unnecessary to adopt a more perfect apparatus, which might 

 easily have been devised had we felt that it was necessary. 



A more serious cause of error may occur in the preceding 

 method ; resulting from the possible impurity of the spores 

 taken from a field of peniciiliiim which has developed in contact 

 with common air. This field receives, every instant, and has 

 received throughout its growth, particles of dust which have 

 fallen from the atmosphere ; thus, it may not be, and, as a 

 matter of fact, is not, free from the germs of other fungoid 

 growths.* 



* Some observations in the preceding chapter enable us to account for 

 the vast number of germs which are constantly falling on the surface of 

 everything. We may here allude to the use we have made of flasks, 

 shaped as in Fig. 17, and holding from 250 c.c. to 300 c.c, which are a 

 third part filled with an organic liquid, and are closed up when boiling. 

 They contain no air when cool, and are opened in series of 10, 20, &c., 

 out of doors, and closed up again immediately. The air rushes violently 

 into the vacuum, and thus we introduce about 200 c.c. of air, with all 

 the particles of dust contained in that air, into each flask. It has been 

 proved that a certain number of these flasks undergo change in the 



