284 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1928 



density of the intruded litliosphere is necessarily increased, and the results out- 

 lined by Barrel! are in full accord with the doctrine of isostasy. 



It would be possible, by rewriting certain portions of the paper, to bring 

 It more nearly into harmony with the most recent data and viewpoints. How- 

 ever, it is thought best to publish the article essentially as Barrell wrote it, with 

 some comment on points that wnll be considered either wrong or questionable 

 in the light of present knowledf,'e. For example, the thickness of the acidic 

 crust assumed by Barrell is too great according to recent findings in more than 

 one field of investigation. After careful analysis of data from studies in 

 •radioactivity Jeffreys states, " Hence the theory enables us to make the quan- 

 titative suggestion that basic rocks predominate in the crust at a depth not 

 greater than 16 kilometers." ^ Later, discussing seismological data, he also 

 concludes " that the thickness of the granitic layer of the continents is of the 



Figure 1. — Diajirammntic section across the North Atlantic, to illustrate theoretic 

 relations of basic intrusions rising from tlie bas;iltic a.sthenosplicre into the orig- 

 inal, more acidic lithosphere, resulting in continental fragmentation in obedience 

 to the forces reestablishing isostacy. Horizontal scale, ca. 480 miles = l inch; 

 vertical scale considerably exaggerated 

 B, Baffin Land; G, Greenland; I, Iceland; F, Faroe Islands; H, the Hebrides; 



S, Scotland and Ireland. 

 There are also indicated — 



1. The surface profile, exaggerated, ocean water omitted. 



2. The granitic or diorltic lithosphere, strongest in its middle part, containing in 

 its upper part lopolithic intrusions of basic rock. 



3. The mean depth of isostatic compensation, the base of the lithosphere. It has 

 in reality no sharp boundary. 



4. Tlie upper part of the asthenosphere, or shell of weakness, penetrated by basic 

 magmas from below and yielding to warping loads in tlie crust above. The origi- 

 nal rock is probably of an intermediate chemical composition, that is, a diorite. 



5. The thick basic shell. The part here shown is solid, like the rest of the earth, 

 but generates molten rock. This tends to rise to a higher level. Since it is a source 

 of magmas, it Is presumably also weak and forms the deeper part of the asthenosphere. 



order of 15 kilometers." ° Joly, from studies in radioactivity in which he 

 considers only the effect of uranium and thorium, concludes that the continental 

 crust is not over 31 kilometers thick, assuming that it is granitic throughout.' 

 Holmes and Lawson,^ estimating the additional effect of potassium, have reduced 

 the figure to 15 kilometers, in close agreement with Jeffreys' i*esult. These 

 estimates are conspicuously smaller than the thickness of 50 to 75 miles as- 

 sumed by Barrell. but the discordance is not necessarily as great as it may 

 appear. Most of the results based on studies in radioactivity assume a shell 

 made of typical granite throughout, with a uniform content of the radioactive 



s Harold Jeffreys, The Earth, p. 87. Cambridge University Press, 1924. 

 8 Idem, p. 179. 



' J. Joly, Surface History of the Earth, pp. 71-78. Oxford University Press, 1925. 

 ^Arthur Holmes and Robert W. Lawson, Potassium and the Heat of the Earth. Nature, 

 May 1, 1026, pp. 620, 621. 



