538 



A CONTRIBUTION TO THE STUDY OF THE PRECIPITINS, 



It will be seen from the results recorded in the Table that the 

 amount of precipitate is directly proportional to the amount of 

 antiserum, i.e., the quantity of precipitin. 



Table viii. 



More accurate methods, however, are reqiiired to determine the 

 I'elation between the weights of interacting antiserum and homo- 

 logous protein and the weight of the precipitate formed. The pre- 

 cipitates have been, therefore, weighed, and an examination of the 

 precipitin i-eaction carried out by gravimetric methods. Measured 

 quantities of fluid antiserum or weighed quantities of dried anti- 

 serum have been mixed with weighed amounts of protein in large 

 tubes for the centrifuge. The tubes have been made up to a 

 fixed volume with saline solution and allowed to stand 48 hours 

 for the interaction to take place. The superfiuid above the pre- 

 cipitate has been removed with a pipette and the precipitate 

 washed five times with saline solution. Each time the precipi- 

 tate has lieen mixed with 50 c.c. saline solution and the precipi- 

 tate separated by spinning in the centrifuge. The precipitate 

 has been then washed five times in the same way with 50 c.c. 

 distilled water. The precipitate has been transferred to suiall 

 glass tubes with thin walls, weighing about 4 gms. These tubes 

 could be spun in a small centrifuge, and in this way the precipi- 

 tate has beeu washed with absolute alcohol and finally with ether 

 free from water. The tubes with their contents have been placed 

 in an oven at 80°C. for several hours and thence put into the 

 desiccator. The tubes have been kept a fixed time in the 

 desiccator and weighed. The weights have been checked three 



