658 APPENDIX: LABORATORY METHODS 



e. Protocols and records. Protocols of each protection test are kept 

 in loose leaf notebooks for a year, after which they are filed in fire- 

 proof drawers. Samples of protocols of Type I and Type II tests ap- 

 pear on pages 659 and 660, and on page 661 there is a sample of the 

 protocol for reporting to the National Institute of Health. The poten- 

 cies of the various lots of serum for distribution are recorded on indi- 

 vidual lot record cards. 



Note: The method of potency testing has been modified to the extent 

 that five dilutions of each serum are used instead of three, although 

 the range of dilutions remains the same ; and the number of mice in- 

 jected with each dilution is six instead of ten. For example, in place 

 of ten mice on each of the dilutions 1 : 1 ,000, 1 : 2,000, and 1 : 4,000 

 there are six mice on each of the dilutions 1 : 1,000, 1 : 1,500, 1 : 2,000, 

 1 : 3,000, and 1 : 4,000. 



Furthermore, the method of interpretation described on page 657 

 has certain inherent faults which occasionally lead to an incorrect re- 

 sult. Therefore, we have been using since March 1939 the method de- 

 scribed by Reed.* A simplification of this method is described by 

 Reed and Muench.f Both depend upon accumulating, with the deaths 

 and survivals at the given dilutions, the deaths at lower serum dilu- 

 tions and the survivals at higher serum dilutions. These methods are 

 probably increased in reliability when a reasonably large number of 

 animals is used and when at least two points fall fairly close to, and 

 on either side of, the 50 per cent end-point. 



E. S. R. 



L. A. B. 

 March, 19U 



* Reed, Lowell J. Chapter II. In Biological effects of radiation. Edited by 

 B. J. Duggar. New York: McGraw-Hill Book Co., Inc., 1936. 



t Reed, Lowell J., and Muench, N. A simple method of estimating fifty per 

 cent endpoints. Am. J. Hyg., 27 (3): 493, May 1938, 



