100 THE AMERICAN MONTHLY [Apr., 



ity of the whole blood, including fibrin and red corpus- 

 cles. 



The question as to whether corpuscles when thus soaked 

 will resume their normal size, and how it can be known 

 that such is the fact, is a very important one but it seems 

 to me it can be settled to a satisfactory degree, by a con- 

 sideration of the physical principles involved. The ex- 

 terior portion of the corpuscle, although not a separate 

 skin or covering, but an integral part of the cell, acts as 

 a dialyzing membrane and allows the passage of fluids 

 through it only in proportion to their density. Hence 

 by the selection of a fluid with a proper specific gravity, 

 we can feel assured not only that the cell will refuse to 

 absorb the fluid and swell indefinitely, but that it will 

 only absorb an amount equivalent to that which it had 

 in the body during life. 



Now before we begin measurement of the corpuscles, 

 let us consider the principles involved and the tools we 

 are to use. There are, as is well known, difl'erences be- 

 tween the measurements of human blood corpuscles, by 

 difi'erent observers, dependent on various causes. 



It is an interesting fact that the tendency of the later 

 work has been to show that the earlier observers made 

 the -measurements too- small. The great weight of au- 

 thority now is in favor of the statement that the average 

 size of a human blood corpuscle is 1-3200 of an inch. For 

 a long time, opinions were divided between 1-3300 and 

 1-3250, then the latter figure was accepted with reserva- 

 tions, then they wavered between 1-3250 and 1-3200 and 

 now there are even some observers whose experience has 

 been that 1-3200 of an inch is rather below than above the 

 limit. My own observations have led me to reach this 

 conclusion. The late Dr. J. B. Treadwell in an elaborate 

 series of researches published in the Cyclopaedia of Medi- 

 cal Sciences makes an average a little over 1-3200. The late 

 Dr. Richardson, whose extensive measurements are well 



