Phifsical Cans/ 1 lid ion o/' Solids cuid Liqni(h, 413 



argument which militates mainly against the hypothesis, is 

 drawn from the phsenomena of expansion and contraction, 

 wiiich are impossible, as it is framed at present, which is ex- 

 cessively nearly, and in most parts precisely, the same with 

 Ur. Hooke's {^ee Micr agraphia). 



In a number of the Annals of Philosophy, (but at what time I 

 cannot exactly state, since I have not the series at hand,) I en- 

 deavoured to show how the particles of solid matter, being al- 

 wavs in contact with each other, and obeying the laws which are 

 known to exist, may alter their relative position, and thereby 

 j)rodiice a change in the volume of the entire mass. The parti- 

 cles being always in contact with each other in certain points, 

 their order of arrangement admits of every variety between the 

 angles of 60° anfl 90°, being held i?i equilibria by the balance 

 of two opposite forces : hence may result every variety of form 

 in crystals, as well as the direction of the cleavages, as also 

 the phaenomenon of the enlargement of the acute, and dimi- 

 nution of the obtuse angles. According to this hypothesis, the 

 force of cohesion is produced by the actual contact of the par- 

 ticles of matter, which force is so greatly diminished by se- 

 jiaration to the least distance, (Newtoni Principia, lib. i. sect. 

 \'2, 13,) that it is commonly said to vanish : the first sepa- 

 ration to even the least distance destroys that force which is 

 properly termed cohesive; and the j)articles are then held to- 

 gether by the force of the whole [)article, as in liquids, as has 

 been shown in the former })apers. 



Were it possible to dejjrive bodies of all their caloric, or 

 to reduce them to the true zero, then the particles must be 

 in the closest possible contact; /. e. the centres of three adja- 

 cent particles must occujjy the angular points of an equilateral 

 triangle: in the utmost state of solid expansion, all the angles 

 become right angles, as in Dr. Hooke's and Dr. WoUaston's 

 figures; between these extremes, the expansion of solids takes 

 place. It is then easy to compute the utmost degree of ex- 

 pansion of which a simple solid, if such exist, is capable, as 

 well as the distance to which the ])articles of a solid must se- 

 j)arate from each other, in order that it may expand or con- 

 tract during fusion. Form a rhombic parallelopipedon of small 

 spheres, placed in rectilinear rows, and so that each sphere of 

 one row shall be in contact with two spheres of tlienext. Let 

 A be one side ot the rhombus; a one of the acute angles; K the 



tabular radius; the solid content is • — -r- — - if there be 



71 spheres on each edge oi" the solid, the diameter of cacli is 



A , . ,. A . ... . /.. A-i . , 



— , and Its radius -, ; its solid content is —r—- ; since there 



u ill 0«' 



