576 BIOLOGY OF PNEUMOCOCCUS 



body available for protection. Felton 402 suggested that an action 

 of this nature might be responsible for the zonal effects mentioned 

 above, or it is possible that Goodner and Horsfall's explanations 

 already cited may be nearer the truth. 



Confusion in the standardization of antipneumococcic serum has 

 been occasioned by the lack of a universally accepted control se- 

 rum. The need was supplied by Cole in 1918, and Felton (1930) 405 

 distributed serum "F 146" to many laboratories as a standard for 

 tests on Type I and II serum. The supply was not large, however, 

 and efforts were made in this country and abroad to substitute 

 other serums comparable in potency to F 146. Unfortunately, the 

 present standard serum for the United States (as well as F 146) 

 is supplied in liquid form and, according to Hartley and Smith 

 (1935), 596 is subject to deterioration. It has been the experience of 

 the authors that loss of protective power of a control serum may 

 introduce major difficulties in standardization. Both F 146 and 

 P 11, the present American standard serum, are for Type I and 

 II titrations. Hartley and Smith have proposed an international 

 standard serum, to be distributed in dry form, for Type I and an- 

 other for Type II. The unit values of these standards are based on 

 Felton's F 146. Hartley and Smith have placed their standard se- 

 rums at the disposal of the Health Organization of the League of 

 Nations. If present methods of standardization are continued in 

 use, it follows that similar standards for types other than I and II 

 will be required. 



Tests of therapeutic effect. One of the objections to the protec- 

 tion test as generally employed in mice has been the fact that it is 

 a prophylactic test rather than a means of estimating the thera- 

 peutic activity of serums. Experience has indicated that the cura- 

 tive value of serums in human beings may be expressed in terms of 

 mouse protective units, yet there has been a surmise that products 

 selected on the basis of experimentally demonstrated therapeutic 

 efficacy might yield better results in treatment. Attempts have 

 therefore been made to devise tests in both rabbits and mice for de- 



