Mr. H. F. Tnlbot on a new Property of Nitre. 147 



twenty, or a greater number of these separate portions, each 

 of which is a perfect crj^stal in itself, but altogether uncon- 

 formable to the remainder. Moreover we shall easily per- 

 ceive that the cracks which took place in the film during its 

 cooling are the boundary lines which separate these diiferent 

 portions ; and that after each crack commences a new crystal, 

 which has no connection whatever mth respect to the position 

 of its axis^ with the preceding one. These successive cry- 

 stalline spaces are however, of course, in the most perfect 

 juxtaposition; and their general appearance and effect may be 

 compared to a sort of mosaic pavement formed of slabs of 

 black and white, or coloured marbles, very irregular in form 

 and outline, but accurately fitting each other. 



It appears then to follow as a consequence from this expe- 

 riment, that the crystallization of melted nitre does not take 

 place in the usual way by the formation of a primitive nucleus 

 and the deposition of layers of molecules upon it; but that we 

 have here a new sort of molecular action brought into evidence, 

 by means of which considerable portions of crystal are formed 

 at once, and pass from the fluid state into that of a crystalline 

 solid with an axis in a determinate direction. We also see 

 that whatever be the nature of the force which determines this 

 direction, it often extends to a considerable distance. But 

 since other portions of the fluid are disposed in solidifying 

 to assume another direction of axis, this different tendency 

 causes the film to crack at the point where the opposite forces 

 counterbalance each other, and to separate into several inde- 

 pendent crystals. 



§. 2. 1 now come to relate another property of nitre equally 

 remarkable with the foregoing; and as these experiments are 

 generally successful, and easily exhibited, I trust they may be 

 considered as some accession to our knowledge of molecular 

 action. 



Let a film of fused nitre be obtained in the manner already 

 mentioned, and then let it be allowed to cool during three or 

 four minutes. The plate of glass should be turned round 

 upon the stage of the microscope until the crystalline film is 

 darkened as accurately as possible. Things being thus ad- 

 justed, let the observer touch the film with the point of a 

 needle, while he is observing it in the microscope. 



He will perceive that the touch immediately produces a lu- 

 minous spot on the dark surface, and this spot will slowly ex- 

 pand itself in all directions like a luminous wave. This is a very 

 curious object, but difficult to describe. Its motion is ex- 

 tremelyirregular, its outline continually shifting and changing, 

 and assuming different colours, until finally in four or five 



U 2 



