j8 ALAN ARTHUR BOYDEN. 



into the cloudiness of the controls. The probability of error in 

 such determinations must have been very high and no doubt 

 some of the discordant results obtained by early investigators 

 were due to this very fact. 



Great steps in advance came with the works of Ascoli (1902) 

 and Fornet and Muller (1908, 1910) in the development of the 

 "ring test" and with Xuttall's quantitative measurements of the 

 amount of precipitate formed (1902). Ascoli suggested that the 

 antiserum be carefully layered under the antigen dilution some- 

 what as the Heller's nitric acid test is performed. Then in 

 positive reactions a definite white stratum of precipitate forms 

 at the zone of contact between antiserum and antigen, the so- 

 called "ring," which is quite obvious and gives the reaction a 

 sharp end point. Fornet and Muller further studied this 

 technique and recommended this ring test strongly because of its 

 greater definiteness. Today the best workers employ this test 

 and indeed there is no longer any excuse for the continued use 

 of the older flocculation test. 



Xuttall published methods for making a quantitative floccu- 

 lation test in 1902. His method consisted in the volumetric 

 measurement of the amount of precipitate formed as a result of 

 mixing a definite amount of an antiserum with a definite amount 

 of a certain antigen dilution. Repeated tests on homologous 

 bloods gave an error of about ten per cent, but sometimes when 

 testing an antiserum with a heterologous blood the error reached 

 fifty per cent. This error seems due partly to the small amounts 

 of precipitate measured and partly to variations in the closeness 

 with which the particles of the precipitate became packed. 



The effects of temperature and of salt concentration on the 

 titer of antisera have also been studied. Though there were 

 some discordant results the influence of temperature was found 

 to be well marked, the rate of precipitation increasing with 

 temperature up to about 40 C. \Yith regard to the effect of 

 salt concentration there is less agreement. In general, however, 

 increasing salt concentration decreases the amount of precipitate. 

 Normal physiological salt solution (.85 to .9 per cent.) is generally 

 used today. Hektoen (1918) recommends a salt solution of 

 double the strength of normal saline for use with chicken anti- 

 he states increases the specificity of the antiserum. 



