80 The Precipitins: Methods 



2'45 : 1000 was present, reaction being entirely prevented by acid 

 4"9 : 1000. Sodium carbonate, added in the proportion "66 : 1000 had 

 no influence, stronger dilutions thereof checked reaction in proportion 

 to their strength. Nevertheless precipitation may occur in dilutions 

 containing 5"3 : 1000, and 10 : 1000 does not quite prevent reaction. 

 In other words, the reaction takes place best in neutral or slightly 

 alkaline fluids, an excess of acid being more inhibitive than an excess 

 of alkali. 



Rostoski (1902, a and h, p. 52) comes to diametrically opposite 

 results. He states that precipitation is impeded by an alkaline, aided 

 by an acid reaction. He acidified with acetic acid or an acid salt 

 (Mononatrium phosphate) or very small amounts of HCl, as little acid 

 is required to prevent reaction. Reaction, he states, takes place to 

 a lesser degree in neutral than in acid fluids. Slightly alkaline solutions 

 (degree of alkalinity 5'0) also react well, but as the alkalinity is 

 increased precipitation is impeded. It is probable that the discordant 

 results of Rostoski are in part attributable to his having worked with 

 opalescent antisera (see page 73). 



It is not impossible therefore that the changes in the alkalinity of 

 the blood in health, but more especially in disease (see von Rigler, 1901) 

 may to a certain extent affect the degree of reaction obtained with 

 precipitins. We propose to pursue the question by means of quanti- 

 tative tests, as purely qualitative tests would scarcely show any marked 

 difference. 



I have referred elsewhere to the abnormally high figures obtained 

 with anti-human serum upon the (concentrated ?) bloods of diseased 

 monkeys. In this connection it is of interest to note an observation by 

 Graham-Smith and Sanger (1903, p. 285) published since the above 

 was written. 



" Our experiments on this subject are only three in number but 

 suggest that important differences may be fouud in diseased blood 

 by means of this test. The following observations were made on 

 sera from tuberculous cattle. The first required 2 c.c. of decinormal 

 caustic soda per 100 c.c. of serum to give a pink tint with phenol- 

 phthalein, and the others l"2o c.c. and '8 c.c. respectively. Also the 

 former required 4*25 c.c. of decinormal caustic soda per 100 c.c. 

 to produce a condition in which the serum was liquid when hot and 

 solid when cold, and the latter 2 c.c. and 1*2 c.c. respectively. As a 

 mean of three estimations in each case these sera produced '0375 c.c, 

 •0328 c.c. and '0244 c.c. of precipitum. 



