322 J. HOWARD MUELLER 



exposure to mild oxidizing or reducirg agents in no way dimin- 

 ished the power of infusion or destroyed its activity. It would 

 seem that if hot alkali and strong acids are avoided, it should be 

 possible to preclude loss of activity while carrjdng out simple 

 precipitations on meat infusion. 



This has been found to be the case, and several different re- 

 agents have been employed, without, however, any marked 

 success in the separation of the constituents of the infusion. 

 Precipitation with alcohol, up to 85 per cent concentration, 

 carried out bj^ preliminary evaporation of the infusion to a small, 

 measured volume, followed by addition of the required amount 

 of 95 per cent alcohol, has in several experiments yielded a pre- 

 cipitate which contains the growth accessory substances, since 

 the test organisms grew on media prepared from the precipitate 

 plus peptone, but not in the absence of peptone. The separa- 

 tion is somewhat more complete if the first precipitate is dis- 

 solved in a little water and reprecipitated. However, in either 

 case, a part both of the growth accessory factors and of the 

 amino acids remains in the alcohol filtrate, and it has not so far 

 been possible to separate them quantitatively in this way. 



Lead hydroxide, lead acetate, mercuric chloride and silver 

 nitrate and baryta have all been tried, none of them with results 

 satisfactory enough to follow up extensively. In general, the 

 filtrates from these reagents will permit growth to some extent 

 without the addition of peptone, and better growth in its pres- 

 ence, which is taken to indicate that the growth accessory frac- 

 tion is not precipitated readily by these metals, while the amino 

 acid fraction, perhaps in the form of peptone, is partially throwm 

 down. By precipitation with tannic acid, followed by removal 

 of the latter with Ba (OH)- + Pb (0H)o in the usual way, very 

 similar results are obtained. 



While none of the methods of chemical separation have given 

 successful results, they have at least shown that it is possible to 

 submit meat infusion to such processes without loss of acti\aty. 

 It has, indeed, been observed that where a single preparation was 

 put through several consecutive precipitations, the activity 

 was gradually diminished or lost, but it is at least as possible that 



