OF HEMOGLOBIN, AND ITS SPECIFICITIES. 73 



Dr. S. Weir Mitchell (Proc. Acad. Nat. Sciences, Philadelphia, 1858, 

 x, Biolog. Dept. 2) found that the color of hemoglobin crystals could be 

 washed out with alcohol and water without injury to their form, and that 

 the crystals may even be redissolved in water and again obtained devoid 

 of color but without change in crystalline type. Preyer (Archiv f. ges. 

 Physiologic, 1868, i, 395) records that hemoglobin crystals are rendered 

 less soluble after standing in dilute alcohol, and that they are converted 

 into pseudomorphs when dried or when in alcohol; Nencki (Archiv f. exp. 

 Path. u. Phar., 1885, xx, 332) states that hemoglobin crystals through the 

 influence of alcohol are converted into an insoluble " parahemoglobin, " 

 which has the same elementary composition as hemoglobin ; Struve 

 (Berichte d. d. chem. Gesel., 1881, xiv, 930) completely deprived hemo- 

 globin crystals of their color by treating and rendering them insoluble with 

 alcohol and water, and without changing their form, and he also found 

 (Jour. f. prakt. Chemie, N. F., 1884, xxix, 304) that fresh blood crystals 

 in strong alcohol became completely insoluble in dilute alcohol. 



Loewy (Zentralb. f. Physiologie, 1899, xm, 449) and Hiifner (Archiv 

 f. Anat. u. Phys., 1901, Supplement, 187) both have determined that 

 alcohol so alters the hemoglobin molecule as to render the readily dis- 

 sociable less readily removed, and therefore render it like methemo- 

 globin. In fact, Hiifner has in recent years insisted upon the importance 

 of avoiding the use of alcohol in the preparation of hemoglobin crystals. 

 Kupffer (Inaug. Dissert., Dorpat, 1884), in experiments with the hemo- 

 globin from the dog, and Kriiger (Zeit. f. Biologie, 1887, xxiv, 47), with 

 crystals from the blood of the horse, have found that with each crystal- 

 lization the absorption coefficient in relation to the spectrum is altered, the 

 absorptive ratio becoming higher and higher, which is the opposite to that 

 which should be expected if recrystallization means merely purification. 



The foregoing facts, together with others which will be found in subse- 

 quent pages, show clearly not only that alcohol is injurious but also that each 

 step in recrystallization means probably the stripping off of extremely un- 

 stable or feebly combined radicals which are normal constituents of the mole- 

 cule and which contribute in giving the molecule its distinctive properties. 



The tenacity with which protein molecules hold impurities has been 

 convincingly shown by Schulz and Zigmondy (Beitrage z. chem. Phys. u. 

 Path., 1902, in, 137), who experienced much difficulty in obtaining egg- 

 albumin free from colloidal substances, and that recrystallization from 5 

 to 7 times was often necessary to obtain a pure substance. While such 

 recrystallization does not affect this protein, according to these observers, 

 it without doubt, as stated, markedly affects the hemoglobin. Abderhalden 

 (Zeit. f. phys. Chemie, 1903, xxvii, 484) states that the hemoglobin of the 

 horse once crystallized may yield as much as 0.62 per cent of glycocoll, 

 the presence of which he attributes to contamination with serum globulin. 

 Since serum globulin yields a little over 3 per cent of glycocoll, there would 

 therefore be about 15 per cent of serum globulin present. He did not find 

 glycocoll after the second crystallization. Even after purification, protein 

 crystals may mechanically take up foreign substances from solution, as has 



