338 SUMMARY AND CONCLUSIONS. 



to form crystals. By altering the osmotic conditions, as by changing the 

 composition of the blood plasma by the introduction of a soluble salt, by 

 dilution with water, etc., the amount of fluid in the corpuscles may be so 

 changed that crystals may develop inside of the corpuscle. It is recognized 

 that the amount of hemoglobin in the corpuscle is much too large in many 

 cases for all to go in solution in the amount of liquid that may be obtained 

 by breaking down the hypothetical union of the hemoglobin and stroma. 

 In such a case the hemoglobin remains amorphous in the corpuscles, or is 

 "colloidal," because it has not enough plasma to make up the amount of 

 plasma of crystallization necessary for the formation of crystals. It can 

 not form a hydrate of definite composition. 



In the ordinary preparation of hemoglobin crystals on slides, according 

 to the method already described, this colloidal form of the hemoglobin is 

 produced in the dense protein ring which forms at the margin of the drops. 

 Upon giving this ring the necessary amount of fluid to form crystals, they 

 are at once produced in the readily crystallizable bloods, the entire ring 

 being rapidly converted into a crystalline mass. 



It seems probable therefore that hemoglobin does not crystallize in the 

 corpuscle because of the osmotic properties of the stroma, which keep the 

 solvent at too low a percentage in the corpuscle to allow of crystallization, 

 while it does not crystallize from the plasma in the living animal because 

 the solution is too dilute or the temperature too high for that dilution. The 

 stroma may have a selective absorption for hemoglobin, which is equivalent 

 to saying that hemoglobin forms a combination with the stroma. 



