﻿350 MARSHALL. 



delicate for the diagnosis of typhoid bacilli, as the bacilli alone are apt 

 to cause deflection. 5 



Furthermore, it has been claimed by Ganghofner and Langner(l-i) 

 that deflection varies according to the amount and concentration of the 

 ingredients and that an excess of hemolytic amboceptor and of com- 

 plement destroys the results. The same writers, although they find the 

 reaction a very sensitive one, conclude that it is not adapted to the 

 purposes of clinical diagnosis owing to the many uncertainties inherent 

 in such a complicated procedure. 



Gengou(16) tested the value of the phenomenon as a means of diag- 

 nosing tubercle bacillus, using a dozen different acid-fast bacilli for com- 

 parison. He concluded that there was nothing specific about the reac- 

 tion, and found that serum from an animal immunized against one acid- 

 fast bacillus gave the deflection test when employed with other species 

 of acid-fast bacilli. 



N"eisser and Sachs (28), replying to Uhlenhuth's objections to their 

 forensic test for human blood, emphasize clearly the necessity of numer- 

 ous controls as a basis for judging the results obtained by this method 

 and they show that Uhlenhuth neglected one control, namely, the testing 

 of the deflecting power of the boiled specimen. They say that it remains 

 for experience to demonstrate whether there are nonspecific deflections 

 which will make the test valueless. The negative results of Uhlenhuth, 

 and of Gengou are possibly due to incompleteness in their technical 

 arrangements, especially to the incompleteness of the controls, and these 

 results must be confirmed before they can be accepted. 



If we summarize this work we may say that the general results form 

 an addition to our knowledge of the mechanism of anti-body reaction, 

 although the mechanism of deflection itself remains obscure. What 

 relation this reaction bears to specific anti-complement action or to the 

 absorption of complement by any albuminous serum, or to mechanical 

 precipitation is not clear. Before pronouncing final judgment upon 

 the results described in this review, we must await the work of others 

 in this field, for the number of factors entering into each experiment 

 increases the chances of error in technique and in judgment to an extent 

 that can only be controlled by an abundance of evidence. This is par- 

 ticularly true of the class of experiments in which the serum is obtained 

 from animals immunized by inoculation with infected organs, for here 

 the amount of infecting material must be slight compared to that of 

 organ material inoculated, and therefore there is great danger lest the 



5 Leuchs. J.: Berl. hlin. WcJmsch. (1907) 44, 68 and 107, has replied to 

 Moreschi's criticism and contends that the deflection method is exceptionally 

 accurate, delicate and specific in diagnosing typhoid fever from other bacteria such 

 as paratyphoid. 



