130 Mr. R. Mallet on the Origin and Mechanism of 



can therefore only state in general terms what are the circum- 

 stances which chiefly determine the size of the prisms, leaving 

 to future observers the determination of the experimental data 

 we need. Good opportunities of observation of various me- 

 tallurgic slags and the temperature at which actual lava ejec- 

 tions lose their plastic or viscous condition and assume suffi- 

 cient rigidity to break by a steadily applied force (transverse or 

 tensional), warrant the author's stating it to be between 900° 

 and 600° F. in almost all basic silicates. This setting tempe- 

 rature, as we may call it, is rather higher in the case of acid 

 silicates, such as are most glasses ; and the temperature at which 

 the rigidity is sufficient to admit of fracture by a blow or im- 

 pulse is in all silicates a good deal higher than that appertain- 

 ing to a steady strain. A familiar example of this we see in 

 sealing-wax, which at common temperatures behaves as a rigid 

 body when broken by a blow, but behaves as a plastic one when 

 exposed to a steady static strain. 



Early in 1858 a smaller crater or bocca on Vesuvius frequently 

 discharged, in volleys or singly, large irregular tails of viscous lava 

 weighing often 300 or 400 pounds. When these larger ones 

 reached the surface of the mountain after their fall, they were ge- 

 nerally too tough to be broken either by their fall or by the blow 

 of a hammer ; they speedily cooled, however, first to a tempe- 

 rature at which they could be shattered by the hammer, and 

 afterwards to one at which they could be broken across by 

 a steady strain very inferior to that which would have broken the 

 same mass when cooled to atmospheric temperature. The supe- 

 rior limit of temperature at which the first stage of rigidity was 

 reached was certainly below that of any " redness w visible in 

 daylight but in complete shadow from the sun's rays. It was 

 therefore probably below 900° F. ; and the temperature of brit- 

 tleness as to static strains appeared to the writer to be reached 

 somewhat above 600° F., or about the melting-point of lead. 

 We may estimate it, then, as probable that the splitting tempera- 

 ture in basalt may be reached somewhere between 900° and 

 600° F., and that while the material is above the superior of 

 these two temperatures splitting by the contractile strains pro- 

 duced by slow cooling cannot occur, but that the effective con- 

 tractile strains will be relieved by the stretching and distor- 

 tion of the still plastic mass. We have no actual data as to 

 the amount of contraction for given decrement of temperature 

 for basalt. We may, however, approximate to it by reference 

 to the author's experimental determination of the coefficient of 

 total contraction between 3700° and 53° F. of iron furnace-slags 

 having a chemical constitution closely resembling the mean of 

 most basalts, as recorded in his paper " On the Nature and 



