276 ANALYTIC ORTHOGRAPHY. 
67. Some alphabetists take credit to themselves if their unlatin pages present a Latin ap- 
pearance through the misemployment of Latin characters (but not letters.) They sprinkle 
them with the unlatin, unitalian, unspanish and unfrench letter ka, pervert goo, made for 
a throat sound, perhaps to represent wh; use the capital “A” in one sense and the small 
letter in another, with perhaps neither in its Latin and Romanic sense; use Teutonic W 
for a Latin sound and its letter, and pervert Greek ¢ and y that they may represent Latin 
sounds already represented. Hence, 
68. If we have no use for certain characters, as those of b, p, f, in some of the American 
languages, we can by no means employ them for new sounds unknown to Latin, to give 
a deceptive Latin page, or to accommodate a frontier printing office. Nevertheless, to use 
the 6 character for an aspirate of m, and the / for English wh, would be trivial errors when 
compared with the perversion of ¢ for English sh. 
69. Dr. Latham’s second rule (English Language, 1841, Chap. 1X.) should be remem- 
bered in forming new characters, and especially in the application of diacritic marks. It 
requires “That sounds within a determined degree of likeness, be represented by signs 
within a determined degree of likeness; whilst sounds beyond a certain degree of likeness, 
be represented by distinct and different signs, and that uniformly.” 
70. Examples of allied letters for allied sounds occur in the Latin ca, UW, UV, BP, PF, Fu. 
Greek has 4 d, and its liquid 4 7; a dot over Arabic 7 makes English z; in Persian arrow- 
head, aleph and ain, and also 7 and a variety of z, are distinguished by the position of a 
wedge on the left or right; Sanscrit from 6 forms (English) w, and p-h from p; but ¢, th, 
d, d-h, ave quite dissimilar. In Thibetian, the affinity is exhibited between 0, p, ph; t, d 
(but not ¢h;) and g, k, (but not kh.) 
71. Welsh had a philosophic alphabet before the invention of printing, for the representa- 
tion of the mutes and their phases, and based upon the Roman letters as follows :— 
bb »>d 4 gay 
bv yp dh (wanting.) 
Dd? Re < cay 
PG fi th K ch 
Ww m Y S ng i 
Here m is acknowledged as a nasal }, n as a nasal d, and perhaps ng as a nasal gay. As- 
piration is indicated by a line, which on the left of the gay character, would have given 
it the sound heard in Belgian, and as this is wanting, the laws of permutation which 
would place it in a word, cannot bring it forward. Hence gafr (Latin capér a goat) be- 
comes dy afr (thy goat) instead of dy ghafr—the analogous form. 
