i 3 o SCIENCE PROGRESS 



their own facets. Twinning can generally be explained as a 

 union of the two crystals along a plane face which is the same 

 in the two crystals, but is reversed in one of them, so that the 

 two crystals grow, as it were, back to back. (Sometimes it is 

 an edge which is reversed.) It is highly probable that twinning 

 is due to the fact that the structures which build up the two 

 crystals can fit together in two positions, either parallel, or the 

 one reversed with regard to the other. 



In the rapid and turbulent growth in a labile solution it 

 may well happen that this reversal is more apt to take place 

 than in the more placid growth in a metastable solution. Hence 

 the beautiful stellate groups and tree-like forms which are more 

 characteristic of rapid than of slow growth. 



Now just as two crystals of any substance may grow on each 

 other whether in parallel position or in twin (reversed) position, 

 because they have the same structure and fit together in these 

 positions, so it may happen that two crystals of different sub- 

 stances may grow together in parallel position for the same 

 reason. This is well known in such groups as that of the 

 alums, where one alum, if introduced into the solution of another 

 alum, may continue to grow as a single crystal, and become 

 enveloped by the latter — as is clearly visible when the two 

 are of different colours. In such a case the two substances 

 are absolutely similar in chemical composition, only differing 

 by the substitution of metals which exhibit the greatest 

 resemblances in chemical behaviour ; and this is no doubt the 

 reason why they possess identical structures. Another still 

 more remarkable case is that of calcite and sodium nitrate. 

 A drop of strong solution of sodium nitrate placed on a clean 

 cleavage surface of calcite at once deposits small rhombohedra 

 which are parallel to one another, and to the rhombohedron 

 on which they are growing. The angles of sodium nitrate 

 are very near to those of calcite, but not identical with them. 

 Further, there is clearly a certain similarity of composition in 

 CaC0 3 and NaNO s . 



In the course of some systematic experiments upon the 

 parallel growths of different substances, carried out in the 

 Mineralogical Laboratory at Oxford, Mr. T. V. Barker investi- 

 gated the question whether sodium nitrate also grows in parallel 

 position upon the other minerals of the calcite group, the carbon- 

 ates of iron, zinc, magnesium, and manganese. He found that 



