642 



MR HENRY BELLY SE BA1LD0N ON 



tongue-tip. The first tends to lose its trill and become almost a mere breathing, as 

 in London and South England generally, the second produces the North English 

 a + r + cons, from e + r + cons, while the Scotch seems to lighten the preceding vowel- 

 sound. 



§ 3 L . 2. E + G or CO rimes with 



a) self. 



way : play, 196, 132. 222, 25. 

 say : away, 114, 21. 



b) ae + g. (see S 23). 



c) e + g. say : tway, 73, 2. 



d) ON. ei. 



wey (weigh) : suey (sway), 246, 104. 

 aye : play, 196, 146. 



e) OF. ai. 



way : verey, 82, 7. pley : gay, 75, 39. 



pley : assay, 110, 205 etc. forwayit : assayit, 110, 

 204. 



f) OF. ei. 



play : pray, 126, 34. pley : array, 196, 13G. 107, 



127. 

 pley : affray, 196, 147. pley : deray, 234, 14. 

 away : affrey, 100, 187. 



g) OF. e or ee. 



pley : lufraye (livery), 196, 140. 



§ 32. An extremely interesting and difficult question demands discussion here, viz., 

 the history of the three sounds, OE. se + g and e + g and ON. ei. Dr Curtis has treated 

 the question at great length and very thoroughly, and has arrived at a conclusion 

 different from that formerly accepted. Dr Gerken [§ 6. 5)], while admitting some force 

 in Dr Curtis's arguments, states the objections which can be raised against his view, and 

 appears to lean to the earlier theory. What we may call the received doctrine repre- 

 sents the line of development thus : — 



OE. ae + g>ai \ 



OE. e + g I . Ui>a>§ 



ON. ei { el j 



Dr Curtis represents it thus : — 



OE. sb + g > ai 

 OE. e + g 

 ON. ei 



I . ei 

 } ei ) 



ei>e 



At the first glance Dr Curtis's theory has the advantage of greater simplicity, and, 

 what is more, it is from the phonetic stand-point much more credible, as it, follows what 

 I have called "the line of least resistance." The other theory labours under the disad- 

 vantage of being circuitous and not following this line, for the change from ai to « is 

 not from the back towards the front, but from front to back, i.e., in the reverse direc- 

 tion to that which it takes in the subsequent stage in passing from a to e (Curtis, § 150). 

 Indeed, the direct transition from a to e seems in itself improbable, so that we should 

 have to insert an intermediate diphthong, ai, which would be very like a reductio ad 

 absurdum of the theory. But we must now examine the objections to Dr Curtis's 

 view as collected together by Dr Gerken (§ 6, 5). See also Luick, Untersuch., 

 § 282, etc. 



The first is " that ai and ei are often, especially in the North, interchanged in writ- 

 ing, and that mostly in favour of ai (see Sweet, HES., § 706 ; Ten Brink, Chaucer's 

 Sprache, § 40)." "According to Dr Curtis's hypothesis," continues Dr Gerken, " one 



