What was the problem? The Berlin Treaty spoke of the provinces inhabited by the Armenians. But the Armenians have become scattered in considerable numbers over the whole extent of Asia Minor. This dispersal is the consequence of comparatively remote historical events. To require the Porte to introduce reforms in the provinces inhabited by the Armenians, and to supervise the carrying out of the new measures, would amount to little less on the part of Europe than to take the whole of Turkey under tutelage. But there might be certain districts in which the Armenians were in a majority, and where they might be able to provide the necessary machinery of government, enjoying a certain measure of local autonomy while remaining subjects of the Sultan. Neither the Armenians themselves nor the British Consuls appear to have furnished satisfactory evidence towards such a solution. What is needed by statesmen who have to deal with Asiatic problems is an intimate knowledge of Asiatic geography. During all the long series of our investigations into the Armenian Question this side of the subject was almost ignored. The Armenian Project of which I have spoken embraced within the area of the proposed province outlying regions which present such dissimilar economical and political problems, that it would have been an act of political madness to endeavour to weld them together under the rule of a mere Governor-General. Our own Consuls, partly, no doubt, owing to the vague character of their instructions, fell into the same error. For instance, in estimating the population of the Armenian provinces, vast outlying districts were included, such as the sanjak of Hakkiari belonging to the vilayet of Van, where the Armenian inhabitants are few and far between, and where the character of the country and people is so wild and intractable that they could with difficulty be controlled from an Armenian centre. The problems that are presented to a Governor on the tableland of Armenia are quite sufficient to absorb his attention and exercise his resources without the addition to his jurisdiction of the mountains of Kurdistan, which, if Russia were mistress of the country, would be constituted into a military Government and subjected to military law.

It must be my endeavour, in proceeding to the statistical aspect of my subject, to avoid, as far as possible with the existing Governmental areas, this lamentable mistake. As in the case of the Russian provinces, I shall adhere as closely as may be feasible to the natural boundaries of the tableland of Armenia, such as they have been determined in the preceding chapter and delineated on the little map which accompanies the political chapter of my first volume. Just as it was necessary in some instances, when dealing with the Russian territory, to overstep the limits of the natural frontier, so I am now compelled by the statistical units at my disposal to diverge at certain points from that established line. Reference to the map of which I have spoken (Vol. I. p. 452) will enable my reader to compare the geographical with the statistical area. The latter is made up of the Governments or divisions of Governments indicated in the following table. Since this statement was compiled the numbers of the Armenians have been reduced by the massacres of 1895. In the vilayet of Erzerum between 2500 and 3000 people were butchered; in the town of Bitlis not less than 800, in that of Kharput 500, and as many as 2800 in Arabkir. Reliable figures are wanting for the losses in human life throughout the country districts of the vilayets of Van, Bitlis and Kharput. But they must have been considerable, and whole villages were wiped out. About 50,000 to 60,000 Armenians fled into Russia from the eastern vilayets. But many of these have already returned, and a few years of settled government would enable this prolific people to make good the deficiencies in their ranks. Later estimates, affected by such special circumstances, would be more misleading than those which I now present.

TABLE III.—Population of the Armenian Tableland in Turkey (about the year 1890)

Moslems.Christians.Others.Total.
Armenians.Greeks.
VILAYET VAN[5]
Town of Van10,00020,00030,000
Merkez-Caza of Van7,00027,00034,000
Other Cazas of Van Sanjak35,22928,64463,873
Total52,22975,644127,873
VILAYET BITLIS[6]
Town of Bitlis and Merkez-Caza27,67316,094342 (Syrian Christians).44,109
Other Cazas Bitlis Sanjak18,59314,306......32,899
Total SanjakBitlis46,26630,400...34277,008
Sanjak Mush—
Town and Caza of Mush21,24635,32856,574
Other Cazas42,57225,87368,445
Total SanjakMush63,81861,201125,019
Sanjak Genjh—
Town and Cazas35,3705,583...40,953
Total of the threeSanjaks145,45497,184342242,980
VILAYET KHARPUT[7]
Sanjak Kharput120,00085,0001334422206,756
Sanjak Dersim62,0008,000......70,000
Total182,00093,0001334422276,756
VILAYET DIARBEKR[8]
Caza Palu45,58015,15060,730
VILAYET ERZERUM[9]
Sanjak Erzerumh—
Town of Erzerum26,55410,434484142238,894
Other Cazas207,26157,3583301797266,746
Total Sanjak233,81567,7928143219305,640
Sanjak Erzinjan155,87931,09124562182191,608
Sanjak Bayazid38,8017,885...56847,254
Total428,495106,76832705969544,502
GrandTotal853,758387,746460467331,252,841

The Moslem population may be divided into Turks and Kurds as follows:—

Turks (Sunni Mohammedan)442,946
Kurds (Sunni Mohammedan and Kizilbash)410,812
Total853,758

It may be interesting to add these figures to those which I have given for the Russian provinces. The population of the country as a whole for the statistical area delimited on the map will be represented by the following figures:—

Armenians906,984
Turks489,931
Kurds479,676
Tartars306,310
Greeks52,367
Russians28,844
Others84,439
Total2,348,551

In the case of the Turkish provinces I have found it a task of the greatest difficulty to arrive at a statistical estimate of the population upon which it might be possible to rely. The results resumed in Table III. are the outcome of a long and laborious investigation pursued in the country itself, in which I was sometimes aided, but more often bewildered, by the lists which I had in my possession, and which have either already been published, or were furnished to me by private friends. In the absence of a census conducted on scientific principles, any figures can only be approximately correct. Two possible sources of information exist which, in the first instance, it is natural to consult. The first are the official lists which are published in the almanacs of each Government, and which profess to give the numbers both of Mohammedans and of Christians inhabiting each caza or administrative sub-division. The second are the books of the diocesan authorities who, under the 14th and 96th Articles of the so-called Armenian constitution (of which I shall speak later on), are enjoined to maintain complete records of all births and deaths among Armenians in the diocese, and to provide copies to the Central Bureau of the Patriarchate in Constantinople. But the diocesan authorities are chary of recording information which conflicts with the number of Armenians who are placed for purposes of taxation upon the Government lists, and these lists themselves are founded upon a system of which it is the tendency to underrate the number of the population, Mohammedan and Christian alike. Owing to the seclusion of women in the East, no serious attempt is made to count the female population; while in the case of males the figures in the official statistics are derived from the military census, which is at best a very imperfect record, and which each man strives his utmost to evade. All Mohammedan males are liable to be enrolled in the army, while the Christians are obliged to pay an annual tax which exempts them from military service, and which is incident at birth. In the case of the sedentary population it is probable that the Christians evade this census to a greater extent than their Mohammedan neighbours; for the budget of a Christian family is immediately menaced by the birth of a male child. On the other hand, there are extensive districts on the southern portion of the tableland in which the Kurdish tribes inhabiting them are in a state approaching independence, and have never been counted at all. The official lists must for these reasons be used with much discrimination and care. In one Government they will be compiled with some measure of completeness; in another they will be defective as regards the Armenians; in yet another as regards the Kurds. In addition to this source of information there are the estimates which have been made in particular districts by private people engaged in business, and who know their own district well. The figures which emanate from the Armenian Patriarchate, and which have found their way into the Blue-books, have evidently been designed to subserve a political purpose, and may be dismissed under a sense of disappointment and disgust.