Professor Tyndall on Germs. 311 



dred and thirty-nine of the flasks operated on were exhibited, and 

 not one of this cloud of witnesses oifers the least countenance to 

 the assertion that liquids within flasks, boiled and hermetically 

 sealed, swarm, subsequently, more or less plentifully with Bacteria 

 and allied organisms. 



The evidence funiished by this mass of experiments, that errors 

 either of preparation or observation have been committed, is, it is 

 submitted, very strong. But to err is human ; and in an inquiry 

 so difticult and fiaught with such momentous issues, it is not 

 error, but the persistence in error by any of us, for dialectic ends, 

 that is to be deprecated. The auther shows by illustrations the 

 risks of error run by himself On Oct. 21 he opened the back- 

 door of a case containing six test-tubes filled with an infusion of 

 turnip which had remained perfectly clear for three weeks, while 

 three days sufficed to crowd six similar tubes exposed to mote- 

 laden air with Bacteria. With a small pipette he took specimens 

 from the pellucid tubes, and placed them under the microscope. 

 One of them yielded a field of Hacterinl life, monstrous in its copi- 

 ousness. For a long time he tried vainly to detect any source of 

 error, and was perfectly prepared to abandon the unvarying in- 

 ference from all the other experiments, and accept the result as a 

 clear exception to what had previously appeared to be a general 

 law. The cause of his perplexity was finally traced to the tiniest 

 speck of an infusion containing Bacteria, which had clung by 

 cjtpillary attraction to the point of one of his pipettes. 



Again, three tubes containing infusions of turnip, hay, and 

 mutton, were boiled on Nov. 2 under a bell-jar containing air so 

 carefully filtered that the most searching examination by a con- 

 centrated beam failed to reveal a particle of floating matter At 

 the present time every one of the tubes is thick with mycelium 

 and covered with mould. Here surely we have a case of spontane- 

 ous generation. Let us look to its history. 



After the air has been expelled from a boiling liquid it is 

 diificult to continue the ebullition without '' bmnping." The 

 hquid remains still for intervals, and then rises with sudden 

 energy. It did so in the case now under coiisiderutiou, and one 

 of the tubes boiled over, the liquid over-sprfa.linu- the nsinous 

 surface in which the bell-jar was imbedded, and on whidi, -loubt- 

 less, germs had fallen. For three weeks the infusions lui.l re- 

 erfectly clear. At tht 



renewino 



: the air of the jar, 



it WIS 



exham 



5ted 



, and retilled bv fresh 



air whicl 



1 had passed through a pli 



.g of 



cott 



on-wool. As the air 







icted b^ 







llspotsofi.eniciUium 



remarkec 



)n the liquid whic 



•h had" 



S>oile. 





er. It was at once 



I that the experim. 



■nt was 





iger 



■ous one, as the enter- 



ing air w 



ould probably deta. 





■ of tlti 



"Sp( 



)res of the penieillium 



and diff-use them in the b. 



41-j:ir. 



This 





, therefore, filled very 



slowly, . 



•o as to render tin 







■ a ri 



dnimum. Next day, 



however. 







bsprvc 





: the bottom of one of 



the three 



■ tubes, namely tha 



t containing 



the 



hay-infusion. It has 



