340 SCIENCE PROGRESS 



Mr. Perrett's system lies between (i) and (2a). He proposes to retain 

 the Roman consonants, with the addition of the Anglo-Saxon symbols for 

 dh and th, but to use a physiological notation for the vowels, based on the 

 pitch of them — strokes drawn upwards, rightwards, and downwards, and 

 at intervening angles, for the vowels, in the order i, e, a, 0, u. The retention 

 of the old consonant-symbols makes the script easily decipherable ; and the 

 vowels are designed upon a logical scientific principle, which is easily mastered. 

 On the other hand, I gather from the examples that the vowel-strokes are 

 intended to be discontinuous from the consonants — which renders a running 

 script impossible, adds greatly to the labour of handwriting, and tends to 

 illegibility. 



^loreover, one asks at once, why should only the vowels be drawn upon 

 scientific principles ? Why not all the letters ? Why not adopt at once 

 a complete physiological system as referred to in (i) above — which should 

 have been designed long ago ? Again, if we retain the Roman consonants, 

 why should we not retain the Roman vowels also, distinguished by one or 

 more of the methods referred to in (2 a, b, c, d) above ? The author objects 

 to diacritics, because he says, " From the typographer's point of view, a 

 diacritic means a new letter." This, in fact, is not the case, because almost 

 every fount of type contains all the vowels marked with the three French 

 accents, and also with the diaeresis, giving, with the five unmarked vowels, 

 no less than twenty-five vowel-symbols altogether ; and I showed long ago 

 that, if used with intelligence, these suffice to give us a complete scientific 

 vowel system (for English at least), without the introduction of a single 

 new letter. On the other hand, as the author states, diacritics require 

 what I call back-hand marks in script ; but so do i and t and the Anglo- 

 Saxon symbol for dh (which the author supports) ; and it is easy to show 

 that arrangements can be made by which a complete diacritical system 

 (giving stress as well as vowel) can be constructed which will actually require 

 fewer back-hand marks than are required by our present spelling. Many 

 good systems under (2c) and {2d) may also be designed ; and there are also 

 mixed methods, and the late Dr. Larrison's system under (2a) — all of which 

 are preferable in some respects to Mr. Perrett's, though they may be inferior 

 to it in others. 



It is certainly good to devise as many systems as possible, for comparison ; 

 but the problem before us is to determine the best one. This requires a 

 very wide and deep survey of all the difficulties and requirements ; and 

 this book hardly attempts to answer the question, but merely insists on the 

 author's particular solution. We can suggest at least half a dozen other 

 solutions which are as good. Of course, the public can always and now 

 adopt one reform — the dropping of useless letters — which would by itself 

 constitute a great advance ; or, by the introduction of an accent without 

 any other change (as pointed out by me in Science Progress for 1913-14, 

 p. 367), it might amend our spelling immensely — and still more so by adopting 

 both these alterations. But even these, simple as they are, would be a cause 

 of much expense in printing, and would leave the final constructive problem 

 unsolved. 



The book is written in an irregular and garrulous manner ; but it is 

 often witty, always amusing, and contains information which will be of 

 interest even to those who have given some study to the subject. 



R. Ross. 



