674 SCIENCE PROGRESS 



really are artefacts. The author has published a large number of highly 

 interesting scientific papers on these particular flints, and also on cognate 

 matters, and he has now published his chief arguments and conclusions, 

 which are of a most far-reaching character, in book form. The greater part 

 of the matter contained in the present volume has been published previously 

 in scattered articles, and the book bears the obvious marks of its origin from 

 separate essays. The chapters do not fit very well into any connected plan, 

 and there is a certain amount of repetition. Moreover, as the book was 

 presumably intended for the educated public generally as well as for professed 

 anthropologists, we think it would have been made more useful and instruc- 

 tive to the former class of reader if a certain amount of preliminary explana- 

 tory and non-controversial matter had been included, perhaps in the shape 

 of an introduction. The book, as a book, is not good ; but any shortcomings 

 which it may have in this respect are quite overshadowed by the importance 

 and extraordinary interest of the subjects with which the separate chapters 

 deal. 



The first chapter deals with flint fracture and flint implements, the second 

 with the oldest flint implements, the third with the relationship of the most 

 ancient flint implements to the later river-drift palaeoliths, the fourth with 

 the ancestry of the Mousterian flint implements, the fifth with " A Rostro- 

 Carinate flint implement," and the sixth (and last) with pre-palaeohthic 

 man in England. There is much in nearly all these chapters which will be 

 instructive even to those who have followed the advances of prehistoric 

 anthropology, and by far the greater part of the work is entirely original ; 

 and indeed the only places where the author does not hold the reader's 

 attention so closely are where he is dealing with discoveries which are not 

 his own, as, for instance, with the Piltdown discovery. It is perhaps about 

 the first chapter that the chief controversy must rage ; for it is in this chapter 

 that the author explains his criteria for distinguishing artificial from fortuitous 

 chipping of flints. And we must say that, so far as this chapter is concerned, 

 the criticism made at the outset does not apply, for the technical terms 

 used are fully explained, and by making a reasonable effort the uninitiated 

 could follow the technical arguments involved. The reader is told the 

 meaning of such terms as " striking-platform," " cone of percussion " (positive 

 and negative), " firaillure," " Ripple-marks," " truncated flake-scars," 

 " opposing cones of percussion," etc., etc. And the author then sets out a 

 detailed and absorbing argument in regard to the criteria of artificial flaking. 

 Thus he finds, firstly, that there is a constant angle in human flaking, and 

 divergent angles in fortuitous flaking ; also that there is a certain squatness 

 in fortuitous flakes ; also that fortuitous flakes tend to go deep into the flint, 

 thus leaving a step at their point of final separation ; also that there are 

 numerous prominent ripple-marks in fortuitous flaking ; and also that 

 truncated flake-scars are more numerous in fortuitous flaking. It is in this 

 discussion that we have the crux of the whole matter. The reader will observe 

 that we have here not so much any absolute criterion (if there had been an 

 absolute criterion the great discussion would not have been so long as it has 

 been) but a series of converging indications or proofs. To follow the argu- 

 ment properly the chapter must be read in full, and probably read twice. But 

 it is perhaps not out of place to point out that in an argument of this particular 

 character the law of probability has to be considered. If, for instance, we 

 take the first criterion, that of a constant angle in the flakes, we must 

 admit that, whilst the majority of flints showing (to take the simplest case 

 of all) two contiguous chips, would not have those chips at a constant angle 

 if the chipping were fortuitous, yet a certain minority of flints, thus fortui- 

 tously chipped, would show the required constant angle. If, for instance, 

 we take the case of two chips, perhaps one out of every eighteen specimens 

 would show the required approximation to the necessary constancy of angle. 



