Ill 



and we aimed at selecting these in such a way thai as 

 many as possihle of the experiments gave results, which 

 were easily measured. Take for instance from the toxin- 

 mixture to be examined the following amounts 1,0, 0,8, 

 0,6, 0,5, 0,4, 0,3, 0,25, 0,2, 0,18, 0,12 and 0,1 cc. and add each 

 separately to 10 cc. emulsion of blood. Put all these 

 test tubes at the same time into the incubator and then 

 in the ice safe. The best result is obtained from this 

 series when a haemolysis of about 10% occurs in the 

 case of the doses 0,40,25 cc. When we know before- 

 hand approximately the amount of toxin with which 

 this occurs, the number of lubes may be diminished, 

 excluding the three first and the three last. In most of 

 the cases, each series included 6 or 8 different doses of 

 toxin. 



It would be well, from the beginning, to emphasize 

 that in these investigations errors of experiment occur 

 much oftener than in ordinary physico-chemical ex- 

 periments with solutions. It may, for instance, happen 

 that a greater amount of toxin has a weaker action 

 than a smaller one. This may be accounted for by the 

 fact that in some cases the added number of blood- 

 corpuscles, through inadequate shaking, was not the same 

 in all the tubes. This occurred in a very high degree 

 in some cases, where the blood corpuscles from some 

 unknown reason showed a strong tendency to aggluti- 

 nate. When the diluted blood had been standing for 

 some time in the test tubes, before toxin was added, a 

 hasty glance was sufficient to show that in this case 

 the number of red corpuscles differed very greatly in 

 the different test tubes, varying in the proportion 1 : 5. 

 This arose from the very quick sedimentation of these 

 agglutinated corpuscles, so that the blood from this 

 horse must be rejected as unsuitable for our experi- 



8 - 



