46 CRYSTALS AND LIVING UNITS 



at a higher temperature, in sulphur-yellow, rhombic tables of the 

 monclinic system. The red crystals turn yellow when heated, and 

 resume their red tint on cooling. The yellow crystals obtained by 

 sublimation retain their colour when cooled ; but, on the slightest 

 rubbing or stirring with a pointed instrument, the part which is 

 touched turns scarlet, and this change of colour extends with a slight 

 motion, as if the mass were alive, throughout the whole group of crystals 

 as far as they adhere together." Then again : — " Nitrate of potassium 

 usually crystallises in the form of arragonite : but if a drop of the 

 aqueous solution of this salt be left to evaporate on a glass plate 

 and the crystallisation observed under the microscope, it will be 

 found that, side by side with the prismatic crystals at the edge 

 of the drop, a number of obtuse rhombohedrons of the calcspar 

 form are produced, just like those in which nitrate of sodium 

 crystallises. As the two kinds of crystals increase in size and 

 approach one another, the rhombohedrons become rounded off 

 and dissolve, because they are more easily soluble than the others, 

 while the arragonite-shaped prisms go on increasing in size. 

 When the two kinds of crystal come into immediate contact, the 

 rhombohedral ones instantly become turbid, acquire an uneven 

 surface, and after a short time throw out prisms from all parts 

 of their surfaces. Contact with foreign bodies also brings about 

 the transformation of the rhombohedrons while they are wet. If 

 the drops are so shallow that the liquid dries round the rhombo- 

 hedrons before they are disturbed, they will remain for weeks 

 without disintegrating, and bear gentle pressure with foreign bodies 

 without alteration ; but stronger pressure, or scratching, or the 

 mere contact of a prismatic crystal of saltpetre, causes them to 

 change." 



These facts, together with those already cited, seem to show 

 clearly enough, not only that the crystaUine form of any crystal- 

 lisable material is variable to a remarkable extent when it is 

 allowed to crystallise under different conditions ; but that, even 

 when formed, a crystal produced under a certain set of conditions 

 may be compelled by its very nature, when these are changed, to 

 undergo an entire molecular rearrangement before a polar equi- 

 librium can be again established between the same molecules and 

 the new influences to Which they are subjected. It will be well 

 that all such facts should be thoroughly borne in mind when we 

 have to consider the transformations that occur, or that may be 

 deemed possible, in simple or elementary forms of living units : for 



