356 SCIENCE PROGRESS 



Society), and j by dzh. Personally, I object to adopting what 

 may be called the India Office alphabet as the final scientific 

 orthography for the rendering of all tongues all over the world ; 

 for the reason that it is not strictly logical, and does not take 

 into account the need for expressing a variety of sounds and 

 combinations of sounds which occur not only in English but in 

 many other important languages. Take, for example, the matter 

 of aspirated letters. In English, and very much so in Arabic 

 and the languages of India and of East Africa, we have aspirated 

 consonants — th, ph, dh, sh, kh, ch, and zh, which require the h to 

 express the aspiration that follows. This need precludes the 

 use of th and dh to express the English th in ' this ' and ' think,' 

 and zh for the French j or the z in ' azure,' or ph in ' physic ' 

 (which last we pronounce literally as p h in ' Clapham ' and 

 * haphazard '). The Arab name of the Muhammadan university 

 at Cairo— Al-Azhar — is pronounced ' Az-har,' and not as if it 

 were written in French, ' Ajar.' A large proportion of the klis> 

 that we meet with in Indian words are not pronounced like the 

 ch in the Scotch ' loch,' but like the aspirated k in ' bloc£/zead.' 



Consequently, we need in our scientific alphabet single 

 symbols for the German and Scottish ch (or the kh so con- 

 stantly used in transcribing Arabic, etc.), for the modern Greek 

 X, for the quite different ch in the English, Spanish, and Indian 

 languages, for the gh represented by the Arabic p, for th in 

 ' theory ' and th in ' that,' for the sh and zh. We require to 

 discriminate between the ordinary s represented by s in ' sea ' 

 and 55 in ' fussy,' and the alveolar Arabic 5 (,j*), between the 

 German ch in ' machen ' and that in ' ich ' and ' dicht.' (This 

 last, represented in the standard alphabet of Lepsius by %, is 

 practically the pronunciation of the Polish £, and the sound is 

 alleged to occur in certain forms of East African speech. It is 

 a transition between the English sh and the German ch — 5 and %.) 

 Then, again, we must provide a symbol for the Arabic d (^>), 

 V(A»), and z (Jo), most of which are alveolar, almost palatal pro- 

 nunciations of the ordinary d, t, and z. The nasal consonant — 

 expressed clumsily by ng in most modern European tongues, 

 and still more clumsily by the apposition of two gutturals in 

 Greek — must have a symbol all to itself, and this is most con- 

 veniently supplied by the n. It is true that n is associated in 

 Spanish with the palatalised n, but this is really nothing but 

 ny % two separate sounds combined. We are, however, used to 



