404 Allen, Crenshaw, and Merwin — 



the transparency of the former, optical tests are the most prac- 

 tical. Yet we arrive at the same general conclusion in both 

 cases : viz, that the product of hydrogen sulphide on acid solu- 

 tions very commonly gives both crystalline forms. 



F. Effect of temperature and acidity on the crystalline 



form of the products. — The two variables of prime importance 



in their effect on the crystalline form of iron disulphide are 



temperature and acid concentration. Experiments were made 



at several temperatures, as follows : 



1. The products at 25°. — In bottles of about 2*5 liters 

 capacity were put a 5 per cent solution of hydrous ferrous 

 sulphate, several grams of powdered sulphur and a variable 

 quantity of sulphuric acid. The bottles were packed in ice 

 and their contents saturated with hydrogen sulphide. Several 

 series of these bottles were tightly stoppered and left to stand 

 at room temperature for periods of many months. Excepting 

 a single instance no precipitate was ever obtained where the 

 initial concentration was as high as 0*03 per cent by weight of 

 sulphuric acid even when the time was lengthened to a year. 

 Precipitation always resulted when the initial acid was as low 

 as 0*015 per cent.* As Table I shows, precipitation started 

 in one instance where the initial acid was 0*059 per cent. 

 This bottle had been used before for the same purpose, so that 

 one might suspect some crystal nuclei had been left behind ; 

 yet in other cases crystals intentionally added had no effect 

 when the original acid was '03 per cent. Whatever the cause, 

 there can be no doubt of the fact, for the data in Table I show 

 a close equivalency between the acid and the sulphide formed. 

 This is not a solubility phenomenon, for again the table shows 

 that the precipitation once started will proceed in a concentra- 

 tion far beyond the above figure. 



A few data on the rate of formation may be of some inter- 

 est. The first two experiments in the table required about one 

 month, the third 80 days. The products in all cases were 

 crystalline powders and microscopic examination indicated 

 that crystallinity was complete. 



The results of the Stokes method were in good accord. In 

 Experiment 3 (Table I), where the final acid concentration was 

 0'23 per cent, the value obtained for (x — a), the dissolved 

 iron, was 10*20, corresponding to 94 per cent of marcasite. 

 Where the final acid was 0'17 per cent, (x — a) was 10*34, 

 equivalent to 96 per cent of marcasite. These figures depend 

 on results obtained with the purest natural marcasite we have 

 yet found, but not all natural marcasites show exactly the 

 same value for (x — a), and farther on we shall give a reason 

 for believing that this product is pure synthetic marcasite. 

 * Wrongly given 0*15 per cent in a previous paper, loc. eit. 



