Sources of Error 75 



precipitation occurs when the same proportions act upon each other in 

 greater dilution, although the bodies combine. A precipitum which 

 had been resuspended in a very large volume of fluid did not become 

 redeposited, whereas it did become redeposited in a smaller volume 

 of fluid. He states further (p. 294) that antiserum added to different 

 concentrations of albumen gave a greater reaction with higher dilutions 

 than with more' concentrated ones. As has already been noted above, 

 he also finds that an excess of precipitable substance checks the 

 reaction, he thinks mechanically, for if the precipitum is collected and 

 resuspended in saline and in concentrated albuminous solutions respec- 

 tively, the precipitum is rapidly redeposited in the first case, not at all 

 in the second. He appears to be somewhat confused as to the inter- 

 pretation of this result. It is scarcely due to viscidity alone, an excess 

 of albuminous substance causes re-solution of the precipitated particles. 

 (See Antiprecipitins, p. 149). 



Rostoski (1902 b, p. 53) found that more precipitation occurred in 

 0'5 /o and 1 / than in 6 % serum albumin solutions, under otherwise 

 similar conditions. This was still clearer in corresponding globulin 

 solutions. Strong dilutions of serum pseudoglobulin hindered precipi- 

 tation more markedly than serum albumin dilutions, and pseudoglobulin 

 hindered the precipitation of serum albumin. He concludes that the 

 globulin exerts an antiprecipitating action, besides having an effect 

 through concentration. He compares the antiprecipitin to antitrypsin 

 and antirennet in blood (Fuld and Spiro, 1900 01), and to the normal 

 antihaemolysin of pseudoglobulin (Pick, 1901). 



Rostoski (p. 58) also found higher dilutions to favour precipitation 

 of egg albumin by its antiserum, for precipitation took place in 

 1'5/ solutions, whereas it was impeded to some extent in 3'5/ 

 solutions. 



In medico-legal work it will be desirable to progressively dilute 

 a suspected blood sample, and to reach a conclusion upon the highest 

 dilution (within limits) which reacts to a given antiserum. In routine 

 work, as I have stated, I have worked with dilutions of usually 1 : 100 

 to 1 : 200. As the dilution increases, the reaction narrows down more 

 and more, the reactions with the highest dilutions being practically 

 specific. Of course, in medico-legal work, the possibility of blood stains 

 being due not to one blood, but to two or more, has to be taken into 

 consideration. On one or two occasions I have found an avian blood 

 react to an anti-mammalian serum. For instance, the blood of a 

 swan (reported in rny paper of vn. 1901), gave a very slight clouding 



