CHEMICAL EQUILIBRIA 137 



calculated figures may be regarded as very satis- 

 factory, if we except two observations (marked with 

 a ?) which do not fit in at all with their surroundings. 

 For all the observations in which physiological salt 

 solution has been added the calculated values are 

 too low, which perhaps is due to a lower solubility 

 of the precipitate in salt solution than in serum. 



The quantity P has a maximum between A = 3 

 and A = 5 ; the calculation indicates the maximum 

 to be 67 at A = 3-75. The maximum depends upon 

 the dilution of the precipitin increasing with the in- 

 creasing addition of diluted sheep serum. 



The said precipitin does not only give a precipi- 

 tate with serum from sheep, but also with serum 

 from related animals such as goats and cattle. In 

 these cases the normal sera contain enough pre- 

 cipitinogen per c.c. to give 40 units of precipitate 

 just as did the normal sheep serum. But the 

 rabbit's serum does not contain more precipitin 

 than is necessary for the formation of 85 (for goat's) 

 and 35 units (for cattle serum) of precipitate, whereas 

 the corresponding figure for the sheep serum is 120. 

 The constant K sinks from 250 for the sheep serum 

 to 1 80 for the goat serum and to 90 for the serum 

 from cattle. These figures give a measure of the 

 relationship of sheep to sheep, which may be taken 

 as unit, as compared with that of sheep to goat 

 (0-72) and for that of sheep to cattle (0-29 and 0-36 : 

 mean value 0-33). 



If we add casein in increasing quantities to 

 lactoserum a precipitin is formed at first which 



