[67] For the times when these different “communities” were formed within the Ottoman state, see Steen de Jehay, _passim_. In brief, the Greek community was organized in 1453 and the Armenian in 1461. The latter was at first supposed to include all subjects who were not Moslem or Greek Orthodox; those who were not Gregorian Armenians were gradually separated off by a process of differentiation which may be said to be active still. With the growth of the spirit of nationalism in the nineteenth century, the Greek Orthodox community has also been divided. [68] Hammer, _Geschichte_, ii. 237. [69] See Jorga, iii. 167 ff., especially 174, 188. [70] Gritti, 24-25. [71] A. Barbarigo, 149-150 (from a summary of his _Relazione_). [72] Barbaro, 315-329, _passim_. [73] Morosini, 263-267, _passim_. [74] Bernardo, 358-364, _passim_. [75] Zane, 389. [76] _Ibid._, 414. [77] Knolles (ed. 1687), 982. [78] Ricaut, 14-15. [79] These _Spahis_ of the Porte are to be distinguished from the body of feudal _Spahis_. See below, pp. 98-105. [80] This is illustrated by the quotations in the last section of Chapter i, above. See also Menavino, 138 (referring to the pages, he says “Tutti sono suoi schiaui & figlioli Christiani”); Ricaut, 14 (all who receive pay or office from the sultan are called _kul_); D’Ohsson, vii. 203 (“Les employés civils de même que les militaires, suivant l’antique usage de l’Orient, sont assimilés aux esclaves du Souverain, et qualifiés de ce nom—coul—dans toutes les pièces publiques”). In D’Ohsson’s time the term had acquired such a general usage as in the English phrase “your obedient servant.” Della Valle, i. 44, speaking of the entry to the Divan, says, “Tutti sono schiavi;” and Ranke, 9, says, “All were slaves.” [81] D’Ohsson, vii. 149, 207. [82] For example, Lodge, _The Close of the Middle Ages_ (1906), 500; Myers, _Medieval and Modern History_ (1905), 165. [83] The _Tractatus_, ch. viii, says simply 20 years and under; Zeno, 128, says above 10 years. Ramberti and Junis Bey (below, pp. 244, 263) mention pages from 8 to 20 years old; Navagero, 49, says between 12 and 15 years, Trevisano, 229, says that they were taken not at the age of 6 or 7 years as formerly, but at 10 or 12 years. Postel, iii. 23, sets between 12 and 14 as the lower limit, and 18 and 20 as the upper limit. Nicolay, 62, says that the pages were from 8 to 20 years of age; Garzoni, 396, says that they ranged from the tenth to the thirteenth year; Ricaut, 74, fixes the age at 10 or 12 years. Too much reliance should not be placed on Trevisano’s statement as to former times, since hearsay evidence as to Turkish affairs is unreliable. Considering the rougher life in earlier times, it is likely that levies would then have been made of older, rather than younger, boys. The presence of young boys among the pages was due to the selection of unusually promising captives. [84] Myers (as above) mentions the two methods of capture and tribute as successive.