Day and Shepherd — Lime-Silica Series of Minerals. 285 



ordinary appearance under the microscope differed from that 

 of the orthosilicate, a study of the index of refraction showed 

 the supposed tricalcic silicate to be a mixture of the orthosili- 

 cate with free lime. Having failed tp obtain a single tricalcic 

 silicate which did not contain free lime, and because every 

 specimen which we examined, including many which had been 

 prepared by others, also showed the orthosilicate to be present, 

 we were forced to conclude that the tricalcic silicate has no real 

 existence. We also tried fusing the tricalcic silicate composi- 

 tion with a flux, but the product was always the orthosilicate 

 of calcium with free lime. 



Although we are anticipating pyrometric studies which fol- 

 low, a single glance at our diagram, iig. 3, will show that we 

 have found and measured the /3 < > a inversion of the ortho- 

 silicate in all the compositions up to 90 per cent lime, which we 

 could never have done with a tricalcic silicate intervening 

 unless the tricalcic silicate be assumed to possess an identical 

 inversion, — which would be rare coincidence. We have, there- 

 fore, complete optical and pyrometric evidence of the persist- 

 ence of the orthosilicate throughout the supposed tricalcic 

 silicate region, and no compound of tricalcic silicate composi- 

 tion can exist there. 



By way of completing the search for hypothetical compounds, 

 we also examined compositions corresponding to the trisilicic 

 acid, but found that the mixtures of lime and silica from which 

 the salt of this acid might be expected to crystallize did not 

 give any new phases. These mixtures showed well developed 

 psendo-wollastonite with the tridymite eutectic, and neither 

 rapid cooling nor crystallization of the glass at low tempera- 

 tures gave any indication whatsoever of the hypothetical com- 

 pound. 



Thermal Apparatus. — The thermal measurements to be 

 detailed in the pages which follow were obtained for the most 

 part in furnaces and by methods which have been described in 

 sufficient detail elsewhere.* The work which has been done 

 with the apparatus since that time has exposed one weakness 

 which will be given detailed consideration at a more appropri- 

 ate time and place, but will be outlined here for the use of 

 others who may have occasion to employ this extremely useful 

 and accurate apparatus. For certain reasons of convenience 

 rather than of necessity, the platinum coils which we have 

 heretofore employed for heating purposes have usually con- 

 tained 10 per cent of iridium. It has now been found that 

 even in a nitrogen atmosphere this iridium sublimes slowly at 

 temperatures above 1200° and enters and contaminates the 

 wires of the thermoelement, if they are exposed, even in the 



*Day and Allen, loc. cit. 



