Frederick Guthrie on the Fracture of Colloids. 27 



§ 5. If the sheet of glass be made very much larger, or the 

 flame smaller and more pointed, another alteration of the 

 crack-figure ensues. The apical point of the previous figures 

 advances into the sheet ; and this is followed by a fan-like 

 cracking of the glass between the apex and the still nearest 

 circumference. In fig. 4, a shows the cracking of a plate 

 of crown glass, 9 inches in diameter, over an air-gas burner ; 

 b is a 5-g-inch-diameter plate similarly treated. In c we have 

 a plate of crown glass, 3 inches in diameter, which was laid 

 on a cloth and heated from above by a fine blowpipe-flame. 

 If we conceive what was before called the apical point to reach 

 the centre, the heat fracture would become approximately the 

 central-pressure fracture, namely radial. 



§ 6. A piece of plate-glass ^inch thick and a little over 

 7 inches in diameter, cracked when heated in the centre over 

 an air-gas burner, as shown in fig. 5 a. A piece of " sheet- 

 glass " (Chance's), 3 inches in diameter, cracked as shown in 

 b. A slab of resin \ inch in thickness and 3^ inches in dia- 

 meter, heated in the centre by a jet of low-pressure steam, gave 

 the fissures shown in c. Square porcelain tiles cracked nearly 

 straight across in one crack. 



§ 7. Pieces of crown glass of various shape were next 

 examined, with the result which declares itself in fig. 6. The 

 pieces were supported at the point marked c, and the flame 

 applied below the point marked f . 



The figure 6 shows that the same general type is preserved. 

 It instructs us that the apical point seeks one of the nearest 

 points of the circumference. 



§ 8. Experiments were next made for the purpose of as- 

 certaining under what circumstances, if at all, a crack could 

 cross a crack. A circular plate of crown glass was cut by the 

 diamond in concentric rings, and the crack was made to pass, 

 by tapping, completely through the thickness of the glass, , 

 around the whole circumference. Such divided glass on being- 

 heated in the centre over an air-gas burner cracked according 

 to the same type as before. Sometimes the heat-crack would 

 run across the diamond-crack, as though the latter had no 

 existence. Sometimes the heat-crack would follow, and, as 

 it were, adopt the diamond-crack, and then break off. In the 

 latter of such cases the inner circle may be suffering a three- 

 crack fracture, while the outer ring exhibits only a two-crack 

 fracture on the converse. Fig. 7 (a) exhibits the former 

 circumstance, fig. 7 (b) shows the influence of a greater 

 number of concentric cracks. 



§ 9. The heating of the central part of a circular plate 



