76 The Precipitins: Methods 



with anti-sheep serum. In such quite exceptional cases I have 

 attributed the result to mammalian bloods having come in contact 

 with avian when the samples were collected, the damp filter-paper 

 strips having been placed side by side, with the result that the one 

 contaminated the other. (See tests with blood mixtures, p. 140.) 



Reverting to what I have said with regard to the grade of dilutions 

 tested, I would add that Aschoff (1902, p. 192), who' cites an earlier 

 paper of mine in this connection, pertinently adds, " Man muss stets den 

 Verdlinnungsgrad der Sera und die Menge des zugesetzten Pracipitins 

 beriicksichtigen, um die specifische Wirkung richtig zu beurtheilen." 



4. Effect of Preservatives on Antisera and Bloods. 



In my paper of vn. 1901 I stated that the addition of chloroform 

 to an antiserum did not impede its giving a reaction. Its use for 

 purposes of preservation has since been recommended by Uhlenhuth, 

 Stockis, Biondi, Robin, Rostoski, and Strube. Ziemke (17, x. 1901) 

 also used chloroform, but seems to find antisera thus preserved to lose 

 in potency, adding, however, that they also did so when stored dry. He 

 goes too far however when he adds, " Gegenwartig ist keine Conservir- 

 ungsmethode im Stande, uns das Serum in frischem Zustande zu 

 ersetzen," for as I have shown elsewhere (p. 123) antisera may at times 

 be preserved for a considerable length of time. As I noted in my paper 

 of 16, xn. 1901 (and the observation still holds), none of these authors 

 "state the fact which I have observed, namely, that it is necessary, 

 before using it, to drive off the chloroform by placing the test-tube in 

 the thermostat. A small quantity of chloroform in a test-serum will 

 frequently cause it to produce a considerable clouding, which may be 

 a source of error, as the clouding is produced in almost any serum to 

 which the test-serum is added." It is therefore necessary to guard 

 against such a possibility of error due to the preservative being present 

 except perhaps in minimal amount. 



For the same reason I have discarded the use of trikresol, and would 

 advise caution in the use of carbolic acid, as recommended by Uhlenhuth, 

 Strube and others. 



In further support of the above statement I will refer to the 

 experiments made in our Laboratory by Graham-Smith and Sanger 

 (1903, p. 285) upon the effects of various volatile antiseptics, as well as 

 such agents as formalin, mercuric perchloride, and copper sulphate, etc. 



