240 /. H. L. VOGT 



assuming (a) that the crystalHzation takes place in a very thin 

 fluid siHcate melt, and (b) that the mineral crystallizes from a 

 solution that is of a percentage of loo, 90, 80, or at least not below 70/ 



I will give some measures of the size of crystals in some smelting 

 products, slags, etc. 



Diopside. — In one of my earlier melting experiments with 

 almost iron-free CaMgSijOe (containing at most 1.5 per cent 

 FeO), thus in a melt of practically 100 per cent of CaMgSijOe, the 

 time of crystallization (at 1336°) of a melt of 16 kg. (16,000 gm.), 

 according to pyrometer measurements, amounted to thirty-one 

 minutes. There resulted, hereby, individuals of a length of 30 mm. 

 (dzc), locally, perhaps, a little longer, and in druses were found 

 freely developed columnar crystals 25 mm. long parallel to c, 

 9-10 nun. parallel to a, and 7 mm. parallel to h. 



In the common iron furnace, slags cast in slag-stone molds 

 about 0.4X0.3X0.25 m. in size, where the crystalHzation according 

 to pyrometer measurements requires about thirty to forty minutes, 

 I found in slags with about 75-80 per cent augite, freely developed 

 crystals of diopsidic augite 18-20 mm. long parallel to c, 9-1 1 mm. 

 parallel to a, and 6-8 mm. parallel to h. 



Pseudowollastonite. — ^A melting experiment performed by me, 

 in melting 3.5 kg. of pure CaSiOs (with a time of crystallization of 

 about fifteen minutes), gave as a result thin tabular crystals 18 mm. 

 long measured parallel to the base, with some individuals a little 

 longer. 



In slags with about 60-70 per cent pseudowollastonite with a 

 period of crystallization of about thirty minutes, thin tabular 

 crystals with a base of a length in one case of 7 mm. and in another 

 case of 8 mm. were freely developed. 



The melilite minerals. — The largest melilite crystals and the 

 largest slag crystals that I have seen, on the whole, I found in 19 13 

 in a basic slag, rich in CaO and with some AI2O3 and FeO, from a 

 remelting process at one of the Mansfielder works. In the druses 

 here appeared tabular crystals (001) (100) with lengths parallel 



'Under otherwise equal conditions, the size of crystals is a function of the quantity 

 of the segregated mineral (see Die Silikat-Schmelzldsungen, II (1904), 164). 



