130 INFRA-RED ABSORPTION SPECTRA. 



does not show the water bands. This would indicate that water of 

 crystallization is not different^ from ordinary water, as mentioned by 

 Konigsberger. 



The entire absence of the water band at 3.0 /a, which in this compound 

 could not be obscured by the carbohydrate band at 3.1 to 3.43 /a, since 

 it is absent, would make it appear that the 3.0 /a band in the alcohols 

 is not due solely to the OH -group. 



Curves b and c of fig. 130 are from Konigsberger (loc. cit.). The 

 former (muscovite, HoO-KzO-SAUOs-GSiOo), has a band at 2.g fi, but 

 since it has no band at 1.5 /x, and since its whole aspect is unlike the 

 water curve, we can not consider the 2.9 ja band due to water. Curve c 

 is for biotite mica (reddish-brown color due to iron oxide), and has 

 even less similarity to the water curve. As a whole the curves show 

 that the H and O which exist in these compounds, and which tinite to 

 form water, on applying heat, are in a different condition from the 

 combined O and H, which exists as " water of crystallization." The 

 mica curves are for polarized light. Aschkinass^ found that mica has 

 bands of metallic reflection (absorption) at 8.32, 9.38, 18.40, and 21.25 fi. 



Selenite ( Crystalline Gypsum). CaSOi-f-2H20. (Fig. 131.) 

 This is one of the most easily obtained minerals for studying water 

 of crystallization. Moreover, it can be split into thin, highly polished 

 folia, the thinness being necessary in order to obtain transparency 

 at 3.0 fi. 



In fig. 131 the curve a is due to Konigsberger (loc. cit.), who used 

 a fluorite prism. It shows the 1.5 /x water band. Curve b shows the 

 water bands at 2.95, 4.55, and 6.0 /a, the latter being obliterated by the 

 increasing opacity of the mineral, which has a large metallic absorp- 

 tion (reflection) band at 8.69 /x. The comparison spectrum of water 

 is given in curve c, which is for an exceedingly thin film of water 

 pressed between fluorite plates, hence the 4.7 /x band is almost absent. 

 The full curve for water is given in fig. 31, which shows curves for 

 films 0.0 1 and 0.05 mm. in thickness, from which it will be seen that 

 the 4.7 fj. band is considerably deeper. By computation it is found that 

 the film of selenite, of 0.126 mm. thickness, contains a layer of water 

 0.059 mm. in thickness. This is indicated by the greater opacity of this 

 curve as compared with the 0.05 mm. curve in fig. 31. Curve a con- 

 tains a laver of water 0.12 mm. in thickness. The band of metallic 



^In Graham (Otto's Lehrbuch der Chemie, up. 173) the statement is made that 

 water loses its properties, just as other bodies do when they enter into chemical 

 compounds. 



