Frederick Guthrie on the Fracture of Colloids. 29 



is the greatest difference of temperature at right angles to 

 the line. 



Remarks. 



§12. About the fracture by mechanical strain it may 

 appear to satisfy many that the lines of fracture are perpen- 

 dicular to that resultant of the pressure which lies in the 

 plane fractured. A tear in a sheet of paper is at right angles 

 to the two opposing pressures, or rather to their resultants at 

 the point yielding. 



What is a crack ? Which are its beginning and end ? In 

 only one of the above-cited experiments can the growth of a 

 crack be followed. In § 11 when a plate heated at the edge 

 has refused to crack while being heated but cracks on cooling, 

 the crack is seen to extend from the edge of the plate inwards, 

 following, generally speaking, a semicircular path, but some- 

 times curiously modified towards the centre of the curve. 



A crack is neither a line of least cohesion nor a line of 



greatest strain. Nor is it a line where - has a series of mini- 

 mum values. The more perfectly elastic a medium is, the 

 more fully does the crack resemble a flash of lightning or 

 wisely laid railway-line, and the more it departs from the 

 river-course or the descent of a globule of mercury down an 

 inclined but undulating surface. Its path is the curve whose 



course is determined by the integnal of - being a minimum. 



s 



The sudden splitting through of the solid aether by the electric 

 discharge furnishes us with figures by no means remotely re- 

 sembling those of the fracture of glass. Even or rather 

 especially the forms of fig. 4 remind us of this. 



As to the typical form in fig. 2, it has been suggested by my 

 brother, Mr. Charles Guthrie, that this form is a compromise 

 between the circular line of fracture along some isothermal 

 line where the difference of temperature is greatest, or rather 

 the difference of expansion is greatest, with the three lines of 

 relief which would be radii at angles of 120° with one 

 another. 



This is a very suggestive hint ; but, for reasons sufficiently 

 apparent from the foregoing, it is insufficient. 



