it remeindz [u]s ov _ghost_. The
French _lanterne_ woz riten _lant-horn_, az if it had been so kalld from
the transparent sheets ov horn that enklozed the leit. The _s_ in _island_
owez its orijin tu a mistaken belief that the w[u]rd iz konekted with
_isle_ (_insula_), hwereaz it iz the A[n]glo-Sakson _ealand_ (Jerman
_eiland_), that iz, water-land. The speli[n] _iland_ woz stil k[u]rent in
Shakspere'z teim. In _aisle_, too, the _s_ iz [u]netimolojikal, though it
iz historikal, az havi[n] been taken over from the Old French _aisle_.

This tendensi tu olter the speli[n] in order tu impart tu a w[u]rd, at all
hazardz, an etimolojikal karakter, beginz even in Latin, hwere _postumus_,
a siuperlativ ov _post_, woz s[u]mteimz riten _posthumus_, az if, hwen
apleid tu a late-born s[u]n, it woz dereivd from _humus_. In I[n]glish,
this fols speli[n] iz retaind in _posthumous_. _Cena_ woz speld bei peopel
who wonted tu show their nolej ov Greek _coena_, az if konekted with {~GREEK SMALL LETTER KAPPA~}{~GREEK SMALL LETTER OMICRON~}{~GREEK SMALL LETTER IOTA~}{~GREEK SMALL LETTER NU~}{~GREEK SMALL LETTER ETA WITH OXIA~},
hwich it iz not.

B[u]t nou let [u]s luk more karefuli intu the far more important
statement, that the I[n]glish la[n]gwej, if riten fonetikali, wud reali
looz its historikal and etimolojikal karakter. The ferst kwestion iz, in
hwot sens kan the prezent speli[n] ov I[n]glish be kalld historikal? We
hav onli tu go bak a veri short way in order tu see the modern [u]pstart
karakter ov hwot iz kalld historikal speli[n]. We nou reit _pleasure_,
_measure_, and _feather_, b[u]t not veri lo[n] ago, in Spenser'z teim,
theze w[u]rdz wer speld _plesure_, _mesure_, _fether_. Tyndale rote
_frute_; the _i_ in _fruit_ iz a mere restorashon ov the French speli[n].
For _debt_, on the kontrari, we feind, b[u]t [t]ree or four h[u]ndred
yearz ago, _dett_. This iz more historikal therefor than _debt_, bekauz in
French, from hwich the w[u]rd woz borowd, the _b_ had disapeard, and it
woz a piurli etimolojikal fansi tu restore it. The _b_ woz leikweiz
re-introdiust in _doubt_, b[u]t the _p_ woz not restored in _tu kount_
(French _compter_, Latin _computare_), hwere _p_ had at least the same
reit az _b_ in _doute_. Th[u]s _receipt_ reziumz the Latin _p_, b[u]t
_deceit_ d[u]z without it. Tu _deign_ keeps the _g_, tu _disdain_ d[u]z
without it. Ther iz an[u]ther _b_ hwich haz a serten historikal air in
s[u]m I[n]glish w[u]rdz, b[u]t hwich woz o