584 EXPERIMENT STATION RECORD. [Vol.40 



pension to make a final serum dilution of 1:500, 1:800. 1:1.000. 1:1.200. 

 1:1,600, and 1:2,000: Known negative and positive serums and the bacterial 

 suspension without scrum are used as controls. The tubes are placed in a 

 water bath at from 37 to 42° for 2 hours. 



It is said that with this technique a reaction up to 1,000 or more may ho ob- 

 tained in from 10 to 20 minutes, while a positive reaction always appears in 2 

 hours. If the reaction is above 1: 1.000 the horse should be kept under Investi- 

 gation. A comparative study of the agglutination, complement-fixation, and 

 mallein tests from dam obtained from rhe diagnosis of 123 horses which proved 

 on autopsy to have glanders, showed a percental value of 35.8 in the agglu- 

 tination test, 24.4 in the compleniont-lixation test, and 12.2 in the mallein test. 

 Three cases w< re identified by the agglutination test alone. 4 by the complement- 

 fixation test, and 15 by the mallein test. 



The author points out that no test can be depended upon alone, as eaeh 



one has its peculiar value in certain stages of the disease, but that if all three 

 tests are used very few eases of glanders can escape detection, it is urged that 

 for the successful eradication of glanders all stables should be under the super- 

 vision of the city healtli department, and that records of the three tests should 

 be kept and some mars of Identification devised (or every horse by which all 

 records can lie compa red. 



The antigen for the complement fixation test for smallpox. <>. Casagkanoi 

 (Ann. I»*t. Pasteur, 82 [1918), Vo. /". /</<. 469 \10).— The preparation of the 

 antigen I'm- the complement-fixation test for smallpox is discussed. 



The method proposed consisted of collecting the fresh vaccine without the use 

 of glycerin, grinding it withoul kaolin or silica, ami centrlfuging it in the pres- 

 ence of physiological serum to obtain a slightly opalescent liquid. If to the 

 filtrate containing the virus alone sterile leucocytes be added, the antigen is 

 said to give very constanl results. 



Chemical changes in tuberculous tissues, (J. T. C.wnwr.it. (Jour. In1c r f. 

 Diseases, ..", (1919), V*0. .' pp. 31-118). — This article contains a review of the 

 literature on the chemical analyses ol normal and pathologic, particularly tu- 

 berculous, tissues of men and of animals, and a report of analyses by the author 

 Of bovine lymph gland and liver tubercles and of fresh normal lymph glands 

 and liver. 



Bromocresol purple and litmus as indicators for the classification cf 

 tubercle bacilli. E. Fbothingham (Jour. Med. Research, .s.'i (1918), No. ..'. pp. 

 158-156). — Preliminary experiments are reported in which the Smith glycerin 

 bouillon curves with human and bovine tubercle bacilli were studied by means 

 of the color changes in bromocresol purple ami in litmus. 



It was found that in general with the use of bromocresol ns an indicator 

 both human and bovine cultures first turn the bouillon more violet or blue, the 

 color remaining fixed with bovine cultures and turning yellow with human cul- 

 tures. With litmus both typos turn the bouillon more or less blue at first and 

 the bovine cultures remain bluish, while In the human cultures the bouillon is 

 turned yellow. 



The author concludes that both indicators may prove valuable in differentiat- 

 ing types of tubercle bacilli. 



The significance of tuberculosis in infants and children with measures for 

 their protection, A. V. Hkss (Jour. Amtr. \l>d. ASSOC, 72 (191f». No. 2. ///> 

 S.1-H8). — Statistics are given of the absolute number of deaths from tubercu- 

 losis at different ages in New York and other cities, which show that the high- 

 est mortality is during the first year of life, following which there is a quiescent 



