T. Mellard Reach — Exfoliation of Gneiss in Brazil. 131 



of temperature and decomposition in the gneissio rock-masses of 

 Brazil, perhaps the most striking is the large scale on which 

 exfoliation takes place, producing, in combination with other 

 agencies and with structure, characteristic topographic rounded 

 forms or bosses and conical peaks. 



Among many illustrations Professor Branner mentions a rounded 

 peak just east of the Corcovado, where the " surface has flaked oflf in 

 great sheets that have slid down the mountain slopes." In the 

 Gavea (plate xii) these scales wrap round the peak, and it is 

 remarked as a noteworthy feature in the exfoliation of peaks that 

 " the flakes came to a feather edge on the downhill side so 

 that they overlap each other like gigantic scales." 



After referring these characteristic scalings to what is no doubt 

 their true cause, expansion and contraction aided by decomposition 

 along fractures. Professor Branner gives a table of the linear 

 expansion of gneiss at several temperatures, taking one part in 

 every 200,000 as its average expansion for one degree Fahrenheit, 

 which shows that a rise of 103° would produce a lengthening 

 of 1-854 inches, and 10° of 048 inch. 



Evidence is brought to show that the maximum variation of 

 temperature in Brazil is 100°, on which Professor Branner remarks: 

 " It must be confessed that these figures are disappointing, for they 

 lead one to suspect, if they do not show, that exfoliation, on account 

 of changes of temperature and whei'e small areas are exposed, is not 

 so active an agency in Brazil as the exfoliated peaks and boulders 

 constantly suggest." 



One of my objects in calling geologists' attention to this excellent 

 paper is to point out that the cumulative action of these recurrent 

 expansions and contractions seems to have been lost sight of. So 

 long ago as 1888 1 gave instances in this Magazine ' showing that 

 the efi"ect of ordinary daily alternations of temperature on terra-cotta 

 copings set in cement was in time to produce permanent lengthening. 

 I have been observing the phenomena in many examples since, and 

 I pointed out some of these copings at Blundellands to Professor 

 Lapworth and other geologists when on an excursion of the British 

 Association at the Liverpool Meeting last year. I find that the 

 lengthening proceeds very slowly from year to year, that it takes 

 place in the summer, and to the greatest extent in the hottest 

 weather, and not appreciably in the wintei'. I attribute it largely 

 to the stresses set up by the greater expansion of the heated top 

 surface over that of the under-surface. The natural tendency of this 

 differential expansion is to hog up the coping, but this being 

 resisted by its weight the underside is strained and lengthened, the 

 forces being directed longitudinally, and the hogging up taking 

 place at a fractured joint or at a ramp, if there be one, in the coping. 



Few people, I fancy, have any conception of the enormous silent 

 force which is developed by even a small rise of temperature. In 

 the case of rolled iron Sir B. Baker estimates it as a stress of a ton 



1 " Effects of Alternations of Temperature on Terra Gotta Copings set in Cement," 

 etc., pp. 26, 27. 



