160 CARNEGIE INSTITUTION OF WASHINGTON. 



that its front edge is parallel to a principal section; (c) that with central 

 illumination special weight be given to the effects along those edges of the 

 grain which trend approximately parallel to the principal section. If these 

 conditions be disregarded the value obtained for the maximum refractive 

 index will be too low, the value for the minimum refractive index too high, 

 while that of any intermediate refractive index, as /3, may be too high or too 

 low. Failure to recognize these factors may lead, especially in the case of 

 strongly birefracting crystal particles, to refractive-index determinations 

 which are appreciably in error. 



(4) The measurement of the freezing-point depression of dilute solutions. L. H. Adams. 



J. Am. Chem. Soc, 37, 481-496 (1915). 



It is a matter of some importance to ascertam with accuracy the relation 

 between the freezing-point and the concentration of a solution, for at the pres- 

 ent time this constitutes the most accurate means of determining the number 

 of moles present in the solution ; and it is desirable to have such observations 

 for dilute aqueous solutions of salts, where comparison of the results with 

 those derived from conductance measurements is possible. Accurate data 

 have hitherto been few in number, for the reason that not enough attention 

 has in general been given to the necessity of determining •with sufficient accuracy 

 both the temperature and the concentration of the solution in equilibrium with 

 the ice. The present paper is a detailed description of a method of measur- 

 ing the freezing-point depression of dilute solutions by the use of which one 

 is enabled to observe the equilibrium temperature with an uncertainty no 

 greater than 0.0001° and to determine the equilibrium concentration vnth 

 commensurate accurac3^ The depression is measured directly by means of 

 a 50-j unction copper-constantan differential thermo-element (which gives 

 2,000 microvolts per degree) and a potentiometer system which enables the 

 electromotive force to be read to 0.1 microvolt. A prerequisite for the 

 attainment of this degree of accuracy is a condition of stationary equilibrium 

 between ice and solution. This is secured by thorough mixing (by means of 

 a specially designed stirrer) of the solution with a large quantity of ice in a 

 vacuum-jacketed vessel completely surrounded by melting ice. The concen- 

 tration of the equilibrium solution is determined by a zero method in which 

 an interferometer is used as a means of comparison. The sensitiveness of 

 this method is about 2 parts of solute per million of water. 



Results are presented for solutions of the non-electrolyte mannite and of 

 the salts potassium nitrate and chloride at concentrations ranging from 0.004 

 to 0.1 molal; consideration of these results shows that the method is suscep- 

 tible of the accuracy claimed for it. The mole-numbers for the two salts, as 

 calculated from the measurements, agree well wdth those derived from con- 

 ductance data at the lowest concentrations, the only region -s^dthin which such 

 agreement is significant, 



(5) The system anorthite-forsterite-silica. Olaf Andersen. Am. J. Sci. ("4), 39, 407-454 



(1915). 



In the study of the system anorthite-forsterite-silica the following soUd 

 phases were observed : anorthite, forsterite, cristobalite, tridymite, clinoensta- 

 tite, and spinel. The paper describes thermal-microscopic studies of artificial 

 mixtures of the compounds CaAl^SioOs, JMg2Si04, and Si02. The thermal 

 data were almost exclusively obtained from quenching experiments and the 

 microscopic examination was undertaken on the powdered quenched products. 



The equilibrium relations of mixtures of anorthite, forsterite, and silica can 

 not in general be explained in terms of a ternary system with these compounds 



