a pure Albumen from Egg-white. 301 



examination shows it to consist wholly of small circular crystals. 

 This being separated and redissolved in water it may be re- 

 crystallized by simply adding ammonium sulphate to the point 

 of commencing precipitation. A product may be then many 

 times recrystallized in the course of a few days. 



After it has been once recrystallized the product exhibits 

 upon further fractional crystallization such absolute constancy of 

 rotatory power and percentage composition as I believe should 

 entitle it to consideration as a chemical individual. We possess 

 at present no better test of individuality in the case of proteids 

 than the rotatory power and percentage content of sulphur. The 

 former varies greatly in albumen of different origin, and, as stated 

 above, its variations are great even among the different albumens 

 found in egg-white. The proportion of sulphur also differs within 

 somewhat wide limits. 



Products prepared by the process described above and frac- 

 tionated by repeated recrystallization show a rotatory power and 

 a percentage of sulphur (determined by Carius' method upon 

 washed alcohol coagula) which in different fractions vary only 

 within the limits of experimental error. 



In five successive fractions obtained from one preparation the 

 specific rotatory power {olD) varied only between — 30°'6G (second 

 fraction) and — 30 o- 72 (fourth fraction), the mean being 30 o- 69 ; 

 and the percentage of sulphur lay between 1*577 (first fraction) 

 and 1'568 (fourth fraction), the mean being 1571. Another 

 preparation was recrystallized six times and the variations in 

 rotation were from — 30°"65 to — 30°'72 (mean 30 o- 7l) and in 

 sulphur from l'57l to 1'576 per cent, (mean 1573). 



The rotatory power has been determined in preparations from 

 eleven different supplies of eggs, the original products being twice 

 or thrice recrystallized. The extreme variations were from — 30 o- 64 

 to -30°-76. 



The yield is good, amounting to about 80 grammes per litre 

 of egg-white or about 50 per cent, of the whole proteid present. 

 The results obtained appear to offer satisfactory evidence of 

 purity, and the product is probably better characterised as an 

 individual than any material yet employed in chemical research 

 upon proteids. 



On the simplest algebraic minimal curves, and the derived 

 real minimal surfaces. By Herbert Richmond, King's College, 

 Cambridge. 



[Read 5 March 1900.] 

 v. Transactions, Vol. xix. Part I. 



