58 J- Lindhard. 



during circulation has not changed in regard to the number of 

 leucocytes it contains, during the period necessary for taking a series 

 of samples. The author does not attempt however to obtain such 

 an artificial blood-mixture, as he considers the task bound with 

 great practical difficulties. 



In an extensive work Kjer-Petersen^ has specially dealt with 

 the methods of counting the leucocytes and the determination of 

 the error; he also uses the acetic acid method. He takes only 1 

 preparation from each pipette, after having filled 10 pipettes immed- 

 iately after one another from the same incision, in a person whose 

 blood he found to be relatively constant in regard to the leucocytes 

 by means of a series of enumerations from day to day. From two 

 series of 10 countings in the case of such a "constant" individual 

 he finds a variation-coefficient of respectively 2*7 and 42; a colleague, 

 who undertook a series of 10 enumerations from the blood of the 

 same person at the request of the author, found the coefficient 36. 



It is evident that this method used by Kjer-Petersen лу111 

 probably give too high values, since in addition to the errors he 

 has also to deal with possible variations. 



Kjer-Petersen evidently knew the method of counting used by 

 Bruhn-Fåhræus but does not seem to have known his work and 

 has not endeavoured either to form a blood-mixture outside the 

 body; in fact, he commits the great mistake of including in his 

 determination of the error a number of countings from the blood 

 of other individuals, reasoning somewhat as follows: The accuracy 

 with which we count is of more importance than that with which 

 we can count. The reasoning is correct; but the presupposition 

 underlying the possibility of learning anything at" all about the 

 matter is, that the blood-mixture examined must be constant, and 

 the blood used by Kjer-Petersen is not that. His determination 

 of the error is therefore more or less accidental. 



Hasselbalch and Heyerdahl" steer clear of the difficulties by 

 determining an individual "standard-coefficient", which for their 

 special purpose is undoubtedly an excellent method. But the varia- 

 tion-coefficient certainly varies from day to day, and the method is 

 therefore scarcely of any use as a normal procedure. 



As, however, it will be of greatest interest to know, how great 

 an exactness can be expected from the method of counting used, as 

 also how to reduce the unavoidable errors to the least possible, the 



1 Om Tælling af hvide Blodlegemer etc. Aarhus 1905. 



'^ Om nogle fj'siske Aarsager til V^ariationer i Mængden af hvide Blodlegemer. 

 Vidensk. Selskabs Oversigter. 1907, No. 5. 



