502 



SCIENCE PROGRESS 



Table II. 



shows that an apparent increase may occur. It must, however, 

 be clearly borne in mind that this apparent increase, though 

 true for the particular experiment, is not a true measure of 

 Immunity as it affects the host, from which the serum is drawn, 

 for in the swirl of the blood-stream such sedimentation cannot 

 occur. It is, in short, an experimental error which should be 

 eliminated if a correct estimate of Immunity is sought for. The 

 point may be investigated as follows : If there be an alteration 

 of appetite with ingestion, equation (1) takes the form : 



fei = (f V 



fm y.u) <}> 



As the variation of appetite is a very slight one, we may use 

 the approximations : 



f m = b 4- cm or f m = b - cm 



And we find 



for f m = b 4- cm 



yn.= y 



b /b 



tT+- I )-'"(T +m -0 



(i-e 



zjm 



m! 



Mean = — ( e b - 1 J 



(Standard deviation) a = e x mean. 



