Determination of the Wave-lengths of the Hydrogen Lines. 33 



Summary. 



The proteid reaction described by Adamkiewicz is not a furfurol 

 reaction, but depends upon the presence of small quantities of an 

 impurity in the acetic acid employed. Some specimens of acetic acid 

 yield no reaction, and all may be deprived of chromogenic power by 

 distillation. 



The substance essential to the reaction is glyoxylic acid. 



Small quantities of glyoxylic acid are produced during the oxidation 

 of acetic acid by hydrogen-peroxide in the presence of ferrous iron. 

 Under the conditions used in this research, part of the glyoxylic acid 

 thus formed is split up, yielding formaldehyde. 



Glyoxylic acid is slowly formed when acetic acid stands in the air, 

 and more rapidly in the presence of ferrous iron and under the influence 

 of direct sunlight. Most specimens of acetic acid contain small 

 amounts of glyoxylic acid as an admixture. 



A dilute aqueous solution of glyoxylic acid, which may be readily 

 prepared by the reduction of oxalic acid with sodium amalgam, forms 

 an admirable test for proteids when used instead of acetic acid under 

 the ordinary conditions of the Adamkiewicz test. 



In carrying out this investigation we have been led to employ 

 extensively the method of oxidation described by H. J. H. Fenton, 

 and as a result we have in some degree trenched upon the systematic 

 study of the oxidation of organic acids which he has in hand. It is 

 with his consent that such of our observations are published. 



The expenses of the research were met by a grant awarded to one of 

 us by the Government Grant Committee of the Eoyal Society. 



" Preliminary Determination of the Wave-lengths of the Hydrogen 

 Lines, derived from Photographs taken at Ovar at the 

 Eclipse of the Sun, 1900, May 28." By F. W. Dyson, M.A., 

 Sec. E.A.S. Communicated by W. H. M. Cheistie, C.B, 

 M.A., F.E.S. Eeceived January 17 —Bead February 7, 1901. 



The spectrum of the "flash" obtained in observations of solar 

 eclipses furnishes a method of determining the wave-lengths of the 

 hydrogen series with great accuracy, as these lines are strongly shown 

 and sharply defined. As the determination of these wave-lengths is 

 somewhat removed from the general subject of eclipse spectroscopy, it 

 seemed suitable for a separate paper. 



The following determination is made from four photographs taken 

 near the beginning of totality at Ovar, at the eclipse of 1900, May 28, 



VOL. LXVIII. D 



