414 



SCIENCE 



[N. S. Vol. LII. No. 1348 



straight lines, so that an actual gravitational 

 field exists. We show that the totality of 

 curved paths completely determines the field. 

 Two Einstein fields which are essentially dis- 

 tinct can never have the same paths. In par- 

 ticular, the paths completely determine the 

 behavior of light. 



IV. Our final theorem is that the light 

 equation determines uniquely the gravita- 

 tional field. In particular, the paths of par- 

 ticles can be predicted from the behavior of 

 light in the field. 



It follows that the gravitational field pro- 

 duced by the sun can be explored either by 

 observations on the orbits of the planets or 

 by observations on the deflection of light 

 rays. li is not necessary to use ioth sets of 

 ohservations. There is, in particular, a con- 

 nection between the deflection of light (1.7" 

 at the sun's limb) and the motion of the 

 perihelion of Mercury (43" x>er century) : 

 either could have been theoretically predicted 

 from the other — but in this fairyland who can 

 lay down a boundary between theory and 

 practise? Edward Kasner 



Columbia IJNrvERsiTT 



THE AMERICAN CHEMICAL SOCIETY. 



(Concluded) 

 Some proteins from the Georgia velvet lean, 

 Stisolobium Deeringianum : C. 0. Johns and H. C. 

 Waterman. The Georgia velvet bean contains 23.6 

 per cent, of protein (N X 6.25). Salt solutions of 

 optimum concentrations (3 percent.) extract about 

 15 per cent, of protein. From such solutions 2 

 globulins, designated the a- and /S-globulins, and 

 an altiumin may be separated, the 2 former by 

 fractionation with ammonium sulfate and the latter 

 by coagulation from extracts from which the 

 globulins have been precipitated by prolonged 

 dialysis. The proteins are sharply distinguished 

 by their different sulfur- and nitrogen-content, by 

 differences in the percentages of the basic amino- 

 aeids as determined by Van Slyke's method, and 

 by the fact that the ^-gloiulin does not give the 

 Hopkins reaction for tryptophane. The latter ob- 

 servation is of particular interest inasmuch as this 

 amino^acid has been found in all seed globulins 

 heretofore tested. The a-globulin. and the albumin 

 from the Georgia velvet bean both contain trypto- 

 phane. 



The defioiency of cystine in proteins of the 

 genits phaseolus: C. O. Johns and A. J. Finks. 

 Nutrition experiments with the proteins of the 

 navy bean, Phaseolus vulgaris, lima bean, Phaseo- 

 lus l%matus, and adsuki bean, Phaseolus angularis, 

 show that they are deficient in cystine. This amino- 

 aoid must be added before they are adequate for 

 normal growth. The proteins of the navy and 

 lima beans must be cooked as well as supplemented 

 with cystine before they are available. Similar 

 experiments are in progress with the mung bean, 

 Phaseolus aureus. 



Studies on Neoarsphenamines : P. A. Kobek. It 

 was shown in a previous paper that arsphenamine 

 made by Ehrlioh 's method contains methylalcohol. 

 It is now shown that neo- and sodium arsphena- 

 mines made heretofore contain about 30—40 per 

 cent, impurities, consisting chiefly of methyl al- 

 cohol, ethyl alcohol, sulphites and ether. Methods 

 were described for the first time, for making 

 sodium arsphenamine, neo^arsphenamine and a sol- 

 uble mono-hydrochloride of arsphenamine base, 

 which are chemically pure and whose arsenic and 

 sulfur content is close to that required by the 

 theory. Another method was described for ma- 

 king the dihydroohloride of arsphenamine base. 



The colorimetric estimation of tyrosine 'by the 

 method of Folin and Denis: Ross Aiken Goetnek 

 and George E. Holm. As the result of a study of 

 the various factors influencing the color intensity 

 of protein hydrolysates to which have been added 

 the phenol reagent of Folin and Denis, according 

 to their directions for the quantitative estimation 

 of tyrosine, we are forced to conclude that: (1) 

 Tyrosine can not be quantitatively estimated in a 

 protein hydrolysate by the use of the phenol reag- 

 ent because (2) Tryptophane, if present, will also 

 produce intense colors with the reagent, the color 

 produced by one milligram being approximately 

 85 per cent, of that produced by tyrosine at an 

 equivalent concentration. (3) Indole and indole 

 derivatives, contrary to the statement of Folin 

 and Denis, react strongly with the phenol reagent 

 to produce the blue color. (4) Ferrous iron, and 

 apparently any other easily oxidizable material, 

 also reacts with the reagent. (5) There is con- 

 siderable evidence that tyrosine and tryptophane 

 are not the only protein constituents which pro- 

 duce blue colors with the, phenol reagent. (6) The 

 amount of color which is developed in a solution is 

 not a linear function of the concentration of the 

 reactive materia, but the color values fall off 



