BLOOD RELATIONSHIP 



307 



though 



cal resemblance to give similar qualitative 

 different quantitative results. 



To determine this we proceed as follows: a rabbit is 

 given intraperitoneal injections of 5-10 c.c. of defibri- 



Fio. 108. Apparatus for quantitative estimation of specific precipitation 

 (Nttttall and Inchley, in "Journal of Hygiene.") 



nated human blood twice weekly for about six weeks, then 

 bled about a week after the last injection, and the deal 

 serum separated from the clotted blood. We thus obtain 

 a reagent which when added to clear human blood serum 

 immediately gives a copious white precipitate. If the 

 antiserum be diluted 1:50 or 1:100 as a standard, so 

 that it shall always be present in uniform quantity in 

 all the tests made, and the serums to be tested for blood 

 relationship given various dilutions 1:100, 1:1000 



