NATURE 



6i 



THURSDAY, NOVEMBER 19, 190S. 



GEOLOGY (U'' CHIWl. 

 Research in China. \"iil. ii. Systematic ("icology. 



By Bailey Willis. Pp. v+133 + v. (Washington: 



Carnegie Institution, 1907.) 

 ] T is difficult to be quite sure for what class of reader 

 this work is intended. The main facts of interest 

 have already appeared in the jjrevious volume, but in 

 this they are discussed from " the point of view of 

 systematic continental history." In the present st.ile 

 of knowledge this might seem rather a hopeless un- 

 dertaking, and the result does not dispel our mis- 

 givings; much of the explanatory matter is elemen- 

 tarv, and much is merely speculative. 



The reader's confidence in the author is somewhat 

 rudely shaken at an early stage, as, for instance, 

 at the close of chapter ii., where, on p. 34, it is 

 first stated that " there is room to question what 

 features e.xisted in Centr.-i! .Asia during the .Sinian 

 period"; six lines further on, owing to the fact 

 that the .Sinian strata consist of limestone, this 

 becomes " it is a fair inference that practically all Asia 

 draining to the Cambro-Ordovician Sea was low ,ipd 

 featureless." The next paragraph, however, begins, 

 " The fact that .\sia at the opening of the Paleozoic 

 era was a featureless continent has important bear- 

 ings." After this bold identification of fact with 

 inference, it becomes necessary to inquire into the 

 credentials of other so-called facts. One of the most 

 interesting results recorded is the discovery of a 

 glacial till in ancient rocks, said to be Cambrian. 

 Convincing proof is given of the glacial character of 

 this deposit ; its age does not seem to be so clearly 

 established. In the present volume we read, 



" The tillite (sic) passes into a greenish shale 

 . . . including characteristic pebbles. . . . This shale 

 conglomerate . . . grades into the overlying lime- 

 stone, the basal layer of a great thickness of Sinian." 



But if we turn to vol. i., Blackwelder informs us 

 that 



" The Xan-t'ou formation [of which the till is the 

 uppermost member] is limited above by an uneven 

 surface, upon which lies a sheet of conglomerate. The 

 matrix of the conglomerate is a greenish argillaceous 

 limestone and the pebbles are like those in the under- 

 lying tillite. The two formations are therefore related 

 by a basal conglomerate, which the till was well 

 calculated to furnish. ..." 



This statement is sufliciently clear, and is accen- 

 tuated by the two diagrams representing Ihc succes- 

 sion of strata given on pp. 264 and 268, in each 

 of which an undulating line is drawn between the 

 conglomerate and the till. What, then, is in fact 

 the relation of the till to the conglomerate? Do thev 

 pass into each other (Willis), or are they separated bv 

 an uneven line (Blackwelder)? But, again, is even 

 the asserted age of the conglomerate a fact, or is it 

 an inference? Turning once more to vol. i., we find 

 on p. 269 that the till 



" lies at the base of the Cambro-Ordovician lime- 

 stone, from which we obtained Lower and Middle 

 Cambri;m fossils within less than 100 miles . . . 



\0. 2'\38, VOL. 7(,| 



of Nan-t'ou. Hence it is highly probable that these 

 gl.acial beds on the Yang-tzi are of early Cambrian 

 age." 



Though the fossils were found only a hundred 

 miles away, we should still have been glad of addi- 

 tional evidence to show that the beds at Nan-t'ou 

 were on one or other of the horizons they indicate. 



It is of interest to note in passing that since the 

 Sinian formation is equivalent to the Cambrian and 

 Ordovician, it almost precisely corresponds to the 

 Cambrian as defined by Sedgwick. 



The summit of this formation is said to be on the 

 horizon of the Trenton or Middle Ordovician ; above 

 this " it passes by transition into shales which are 

 probably of Silurian or Devonian age." The sug- 

 gested passage by transilio)! of Middle Ordovician 

 into either Silurian or Devonian shales leads to the 

 suspicion that the author uses this term in some 

 esoteric sense. 



The treatment of the Angara and Gondwana beds 

 is one of the most unsatisfactory chapters in the 

 volume ; both series are included under the head of 

 Permo-Triassic strata, while so far as existing observa- 

 tions go, the Gondwana beds begin with the Lower 

 or Middle Carboniferous.' while the most trustworthy 

 evidence we possess points to a Jurassic age for the 

 Angara. There is an inexactitude also in defining 

 the limits of the Angara beds; they are not confined 

 to the northern region indicated by the author, but 

 extend to Afghanistan and through Turkestan, as 

 Musketow has already pointed out. 



Students of the geology of India will be surprised 

 to learn that " no distinctly sedimentary pre-Cambrian 

 rocks are known there," i.e. in the peninsula (p. 23), 

 and those who have given attention to ripple marks 

 will scarcely admit that they are to be taken offhand 

 as evidence of '' waves " in the common sense of this 

 term (p. 38). 



The structural trend-lines of .\sia seem to be drawn 

 for the most part after the maps of von Richthofen, 

 Suess, Neumayr, and Futterer; it is pleasing to find 

 that the generalisations of these great masters have 

 been almost entirely confirmed by recent investigators. 

 But we see no evidence for the postulated " Isle of 

 Tibet." .\ region of which the geology is almost 

 unknown naturally oft'ers great temptations to the 

 theorist. 



There is some internal evidence of hasty writing, 

 such as inaccuracies in translation conveying a rather 

 different sense from that of the original; as an 

 example we may cite the last sentence of the first 

 paragraph (p. 69) from Suess: — 



"The great height of the ranges is accompanied, 

 however, by a relatively even more striking altitude 

 of the valleys, a circumstance which diminishes the 

 differences of level in the interior of the mountainous 

 regions, but the observer is even thus below the limit 

 of eternal snow." 



Tliis is rendered from the French, which runs : — 

 " L'altitude plus forte des chaines a pour contic- 

 partie unc hauteur plus grande des fonds de vall(-es, 

 ce qui attdnue les differences de niveau h I'int^rieur 



1 Neumayr and W.-iaaen roncluded in favour of an Upper Carbonifeniis 

 .Tge lung ago, and Hayden has since shown ihat they must lie even lower. 



