34 Mr Skinner, Observations on the Minute Structure 
simpler than those obtained from glacier ice a description of 
them will now be given. 
In Figure 1 a cast of a portion of a column of ice from the 
Schafloch is shown. Its greatest length is 195 centimetres and 
the direction of this length was vertical in the cave. As the cast 
is a negative all the cavities in the cast are really prominences in 
the ice 1 . The general appearance has given rise to the terms 
‘prismatic structure’ used by Browne 2 and Bonney 3 , and ‘honey- 
comb ice ’ used by Lohmann. The crystals lie side by side with 
their optic axes parallel and at right angles to the surface exposed 
to the source of cold, just as in pond-ice the axes are normal to 
the surface of the water. They, therefore, present to a spectator 
the appearance of an irregular mosaic, like the tops of the columns 
in the Giant’s Causeway. When crystals are formed in this 
manner, if the water contain any impurities, these will be collected 
on the outsides of the crystals where they are in contact with one 
another. As is well known water which contains salts or gases in 
solution has a lower apparent freezing-point than pure water. 
We may imagine the freezing of such water to occur thus : from 
some point six rays spring, forming a star, and on these secondary 
rays grow. It is pure ice which is forming, and the solution of 
impurities, which is becoming more concentrated, is thrust away 
in front. The impure layers, therefore, will solidify in the inter- 
stitial spaces last, and they will on fusion commence to melt 
first. 
If heat be supplied to ice formed in this manner so slowly, 
that the distribution of temperature is almost uniform, the fusion 
of the interstitial matter will occur before the purer ice and the 
individual crystals will be separated sometimes wholly, and at 
other times incompletely, by furrows. This kind of melting is 
especially favoured in ice-caverns where the temperature during 
a great part of the year is only a fraction of a degree above zero. 
The cast (Fig. 1) is that of a case of this kind. The plaster of 
Paris has flowed into the interstitial spaces to different depths up 
to 5 millimetres. The same phenomenon is shown by artificial ice 
if the block be allowed to warm very slowly. It is also shown by 
many alloys on fusion. 
On the exposed surface of many crystals in the Schafloch 
a special kind of marking may be seen. The hexagonal base is 
traversed by a number of lines which are approximately parallel. 
1 A positive may readily be made by pressing putty on the cast. But to get the 
visual appearance of a positive this is unnecessary, as the figure when looked at in 
a certain way reverses itself and appears to be a positive. I find this optical reversal 
especially easy with figure 2 when the figure is inverted. Viewed in a certain way 
a coin which is actually in relief appears to be incised. 
2 Browne, G. F., Ice-Caves of France and Switzerland, 1865. 
3 Bonney, Froc. Camb. Phil. Soc., March 4, 1867. 
