before the Geographical Society, May 27, 1867. 21 
that in early times, say from 600 years before the Christian era 
to 500 or 600 years after it, both the river Oxus and Jaxartes 
flowed into the Caspian, the Aral being non-existent. That 
afterwards, and up to the year 1300, they fell into the Aral, 
and that for the next two hundred years (1300 to 1500) they 
came back into the Caspian, subsequently flowing gradually 
back into the Aral and forming the sea as we now know it, 
Although I know that my colleague will admit that my geo- 
logical data must have some weight, I have to claim his indul- 
gence for venturing to question the views of so eminent a 
scholar respecting the changes of physical features in this re- 
gion that may have happened in the days of history. Sup- 
ported, however, as I am by the opinions of men ‘on whose 
knowledge I place great reliance, I must say that I cannot 
regard the Persian manuscript, which was presented to Si 
enry by a clever chief of Herat, to be a document of sufficient 
value to override the conclusions at which I have arrived on 
many independent grounds. 
Concerning the ancient course of the Oxus, I see no reason 
to differ from the Persian writer and Sir Henry. But when it 
is stated that in the year A. D, 1417 the Jaxartes had deviated 
from its former course and instead of flowing into the Caspian 
(as the ancients had it,) joined the Oxus, and thus, the two 
rivers occupying one and the same bed, came into that sea, I 
must withhold my assent, This is a novel and striking state- 
ment, and before we attach credence to it we must have some 
physical evidence to sustain it. In my state of scepticism re- — 
arding the value of this Persian manuscript, now for the 
t time produced, that which strikes me @ priori as a sign 
of its invalidity, is, that when this region was open to knowl- 
edge through the long-enduring reign of the civilized and lite- 
rary Arabians (say from the 7th to the 13th century), the Aral 
was known and-laid down as a distinct water-basin under the 
name of Sea of Khwarezm. On the other hand, when after 
that period knowledge became dim and local, and civilization 
was at its lowest ebb, then it was that the Aral disappeared. 
My conclusion from this coincidence of the supposed emptyin 
of the Aral, with the absence of records respecting it, mae 
be that the sea had existed during all that time, but that 
there were then no geographers to record the fact, 
In treating this subject, let us first consider the separation 
of the Aral from the Caspian as originally dependent on geo-. 
logical changes of the surface, and then proceed to estimate 
the value we are to attach to the writings of the classical au- 
thorities in reference to a region so very imperfectly known to on 
them. Asa geologist who has studied this Aralo-Caspian ques- 
