ndian records of any sort,
with a definite date.

One of the few cases where the phenomena of _rate_ have been studied
with due attention, is in the evolution of the three languages of
Denmark, Norway, and Sweden out of the Icelandic. What does this tell
us? The last has altered so slowly that a modern Icelander can read the
oldest works of his language. In Sweden, however, the speech _has_
altered. So it has in Denmark; whilst both these languages are
unintelligible to the Icelander, and _vice versa_. As to their
respective changes, Petersen shows that the Danish was always about a
hundred years forwarder than the Swedish, having attained that point at
(say) 1200, which the Swedish did not reach till 1300. Both, however,
changed; and that, at a uniform rate; the Danish having, as it were, the
start of a century. The Norwegian, however, comported itself
differently. Until the Reformation it hardly changed at all; less than
the stationary Icelandic itself. Fifty years, however, of sudden and
rapid transformation brought it, at once, to the stage which the Danish
had been three hundred years in reaching. How many times must the
observation of such phenomena be multiplied before we can strike an
average as to the rate of change in languages, creeds, and polities?

Again--it is by no means certain that the Institutes and the Vedas
represent a contemporary state of things. All doctrinal writings contain
something appertaining to a period older than that of their composition.

Lastly,--the proof that all the writings in question belong to the same
linear series, and represent the growth of _the same phenomena in the
same place_ is deficient. The AEgyptologist believes that contemporary
kings are mistaken for successive ones; the philologist, that difference
of dialects simulates a difference of age. Doubts of a more specific
nature dawn upon us when we attempt to realize the alphabet in which an
Indian MS. of even only eight hundred years B.C., was written. No Indian
MS. is fifteen hundred years old; no inscription older than Alexander's
time. Nevertheless,--though I write upon this subject with
diffidence--the Devanagari characters of the Sanskrit MSS. can be
deduced from the alphabet of the inscriptions; whilst these inscriptions
themselves approach the alphabets of the Semitic character in proportion
to their antiquity: so that the oldest alphabet of the Vedas is
referable to that of the inscriptions, and that of the ins