Bacteria in Organic Infusions. 121 



that in a closed tube or bottle, after such a cloud (of Bacteria) 

 has developed, the Bacteria at a certain period cease to multiply 

 and settle down as a fine powder, leaving the fluid again clear. 

 Such precipitated Bacteria remain unchanged in the fluid for a 

 long period (weeks certainly, perhaps years), and can be readily 

 shaken up and at once recognized by microscopic examination; 

 they are, moreover, not destroyed by boiling : hence it is not pos- 

 sible to miss the detection of a development of Bacteria in a limpid 

 turnip-infusion, examined daily for three weeks or more by the 

 naked eye, and finally, after agitation, by means of the microscope. 



Seeies a. Nov. 23rd. E.iyeriments v:ith Hay-Infusion. — An 

 infusion was prepared by pouring water of about 90^ C. on to 

 chopped hay. The infusion was of a dark sherry colour ; reaction 

 slightly acid. The glass tubes used in this and subsequent ex- 

 periments were about five inches in length, of half-inch bore, 

 rounded at one end and drawn out to a capillary orifice at the 

 other. The infusion in these and subsequent experiments was in- 

 troduced by heating the tube and plunging its capillary beak 

 beneath the surface of the experimental Hquid during the cooling of 

 the expanded air, until the tube was about one-third or half fiUed. 

 Tubes 1, 2, 3 were half filled with the hay-infusion previously 

 filtered, the hquid was boiled in the tube, and the capillary beak 

 fused, as nearly as possible, during ebullition.* 



Tubes 4, 5 were similarly treated, with the difierence that a 

 small quantity of cheese, in a very fine state of division, had been 

 added to this portion of the hay-infusion before its introduction into 

 the tubes. 



Tubes 6, 7. Quantity and character of the infusion as in 1, 2, 3, 

 but the tubes sealed without previous ebullition. 



Tube 8. Quantity and character of the infusion as in 4 and 5, but 

 the tube sealed without previous ebuUition. 



Tubes 9, 10, 11. Quantity and character of the infusion as in 

 1, 2, 3, but rendered slightly alkaline with KHO. Sealed approxi- 

 mately during ebullition. 



All these tubes (1 to 11) were after closure completely sub- 

 merged in boihng water for fifteen minutes, and were then pre- 

 served in a hot-air bath, varying in temperature from 30^ C. to 

 35° C. 



Microscopic and Naked-Eye Apjoearances of the Hay-Infusion 



* The tubes were sealed at the moment of removal from the flame over which 

 they had been boiling. In every case a subsequent recurrence of ebullilion was 

 observed during the cooling of' the upper part of the tube. Dr. Eoberts, of 

 Manchester, has suggested that the occurrence of Bacteria in tubes thus sealed 

 may be explained by their indraught, together with a certain amount of air, at 

 the moment of closure ; but the esperiments of Sanderson, recently confirmed 

 by Cohn, have shown that contamination of fluids by Bacteria only takes place 

 throusrh the medium of impure surfaces or liquids. 



L 2 



