The Several Forms of Calcium Carbonate. 485 



Required 

 I II for CaC0 3 



Loss on heating 43-9 44*2 440 



CaO from precipitation as oxalate . . 56 f l 56"2 56*0 



C0 2 by direct absorption _ 43-2* 44-0 



There is thus no doubt that it is really CaC0 3 ; its density f 

 proved to be 2"54:, but this value may be slightly low because 

 it was determined on material which was partly fibrous or 

 platy. This, together with its optical properties, suffices to 

 characterize it as a distinct modification of CaC0 3 . 



/x-CaC0 3 varies greatly in appearance in different prepara- 

 tions. The smallest grains distinguishable by the microscope 

 are hexagonal plates with faces either plane or with six radiat- 

 ing ribs. When a diameter of about O01 mra is attained the 

 plates usually cease to grow as units and become the nuclei of 

 larger plates or aggregates which reach a maximum observed 

 size of 0*07 mm . These larger plates or aggregates are almost 

 always thickened in the middle by either a parallel growth of 

 smaller plates, a thickening of the radiating ribs, or overlap- 

 ping of plates that are obliquely radiating. In the latter case 

 the aggregates, though still roughly hexagonal in cross section, 

 become lens shaped;}: or broadly elliptical in vertical section. 

 The radiating plates are often found arranged in these aggre- 

 gates so that the opposite triangular segments of the aggregates 

 seen flatwise extinguish together, and give an apparently biax- 

 ial interference figure, 2E = 35°-40°, with the plane of the 

 optic axes parallel to the external sides of the triangles. But 

 aggregates which do not show this regularity of extinction and 

 the plates and star-shaped forms give a uniaxial positive inter- 

 ference figure. This form of calcium carbonate, therefore, is 

 hexagonal. The refractive indices of well-formed plates were 

 found to be 1*550 and 1*650, within about dtO'005 ; the plates 

 were held edgewise by mounting them in viscous mixtures of 

 rosin and piperine of standard refraction. When measured on 

 lens-shaped aggregates both apparent refractive indices are 

 about 0*01 less than the true value ; analysis of such aggre- 

 gates showed that a little water is enmeshed among the non- 

 parallel plates. 



/*-CaC0 3 , when dry, persists indefinitely at room tempera- 

 ture, as far as we can judge ; when heated alone, it goes over 

 into calcite, both plates and aggregates commonly producing a 

 single calcite grain. In presence of water at ordinary temper- 



* The experimental errors all tend to a low result, especially when one is 

 working, as here, with a small weight of substance. 



fAs determined by the pyknometer described by Johnston and Adams, 

 J. Am. Chem. Soc, xxxiv, 567, 1912. 



% The aggregates sketched by H. Vater (Z. Kryst., xxvii, 486, 1897) were 

 probably this form of calcium carbonate. 



