Review of Recent Geological Literature. 183 
magma is considered to need confirmation. In any event this theory 
can evidently not be appealed to as explaining differentiation in a tan- 
gential direction on the earth's surface. 
As to the theory of more general application that any given magma 
will differentiate while remaining in the liquid state into two tr more 
magmas of varying composition, the reader is referred to Becker's clas- 
sic articles,* which present serious objections based on careful experi- 
mentation. The author then presents objections of a more general char- 
acter, but lack of space forbids following the argument further. It may 
be well, however, to quote the conclusions reached: — ■ 
"i. The heterogeneity of the globe is primordial both in the tan- 
gential and radial direction ; it has not undergone very considerable 
modification during geological time. Crystallisation and the slow and 
limited movements caused by the difference in density between crys- 
tals and magmas containing them are the only efficient causes of sensi- 
ble displacements" of materials in magmas. 
"2. The theory of differentiation taken in its full extension, with the 
meaning that modern petrographers have attributed to it, meets with 
gj-ave objections, not to say impossibilities." 
"3. It is either based upon conceptions entirely gratuitous, or the 
facts of detail cited in its support are uncertain or susceptible of other 
interpretations which are more probably correct." a. x. w. 
Les Roches Alcalines caracterisant la Province Pcfrographiqnc d' Am- 
pasindava, par A. Lacroix. (Nouvelles Archives du Museum d'- 
Histoire Naturelle, 4th Series, Vol. I. 4to, 214 pages, 10 plates and 
45 figures, Paris, 1902.) 
The first two chapters of this memoir, embracing all the descriptive 
matter, were briefly reviewed in the November number of the Geol- 
ogist (Vol. XXX, p. 328). The third and final chapter (second fasci- 
cule) has now appeared; it is devoted to a summary statement of re- 
sults, and a discussion of the conclusions of general interest to be 
drawn from the study — as previously stated the subject of the memoir 
is a new petrographic province situated in the northwestern part of 
Madagascar — this province is characterized by the abundance of erup- 
tive rocks which are of many types, representing all the alkaline rock 
families now known. In known geographic area the province includes 
only a small fraction of the total area of Madagascar covering nearly the 
whole of the extreme northern end. However, Lecroix suggests that 
this may be merely an outlier of a possible petrographic province of 
immense size and similar characteristics extending all the way from 
the Transvaal along the eastern coast of Africa through Zanzibar, 
British and German East Africa, and Somaliland, across the Red sea to 
Yemen in Arabia. 
The following table is a summary of the author's conclusions in 
the most condensed form possible. It is impossible in a brief review 
of this character even to enumerate the reasons given by the author 
* Am.Joar. Set., 1897, iii, p. 21 ; iv, 257. 
