A Chinese account of the expedition of Hulaku will be found in Remusat’s Nouveaux Melanges (I.), and in Pauthier’s Introduction. (Q.  R. 115-219, esp. 213; Ilch. vol. i.; J.  A. S. B. VI. 842 seqq.) [A new and complete translation has been given by Dr. E. Bretschneider, Med.  Res. I. 112 seqq.—­H.  C.]

There is some account of the rock of Alamut and its exceedingly slender traces of occupancy, by Colonel Monteith, in J.  R. G. S. III. 15, and again by Sir Justin Sheil in vol. viii. p. 431.  There does not seem to be any specific authority for assigning the Paradise of the Shaikh to Alamut; and it is at least worthy of note that another of the castles of the Mulahidah, destroyed by Hulaku, was called Firdus, i.e.  Paradise.  In any case, I see no reason to suppose that Polo visited Alamut, which would have been quite out of the road that he is following.

It is possible that “the Castle,” to which he alludes at the beginning of next chapter, and which set him off upon this digression, was Girdkuh.[1] It has not, as far as I know, been identified by modern travellers, but it stood within 10 or 12 miles of Damghan (to the west or north-west).  It is probably the Tigado of Hayton, of which he thus speaks:  “The Assassins had an impregnable castle called Tigado, which was furnished with all necessaries, and was so strong that it had no fear of attack on any side.  Howbeit, Halooen commanded a certain captain of his that he should take 10,000 Tartars who had been left in garrison in Persia, and with them lay siege to the said castle, and not leave it till he had taken it.  Wherefore the said Tartars continued besieging it for seven whole years, winter and summer, without being able to take it.  At last the Assassins surrendered, from sheer want of clothing, but not of victuals or other necessaries.”  So Ramusio; other copies read “27 years.”  In any case it corroborates the fact that Girdkuh was said to have held out for an extraordinary length of time.  If Rashiduddin is right in naming 1270 as the date of surrender, this would be quite a recent event when the Polo party passed, and draw special attention to the spot. (J.  As. ser.  IV. tom. xiii. 48; Ilch. I. 93, 104, 274; Q.  R. p. 278; Ritter, VIII. 336.) A note which I have from Djihan Numa (I. 259) connects Girdkuh with a district called Chinar.  This may be a clue to the term Arbre Sec; but there are difficulties.

[1] [Ghirdkuh means “round mountain”; it was in the district of Kumis,
    three parasangs west of Damghan.  Under the year 1257, the Yuean shi
    mentions the taking of the fortress of Ghi-rh-du-kie by
    K’ie-di-bu-hua. (Bretschneider, Med.  Res. I. p. 122;
    II. 110.)—­H.  C.]

CHAPTER XXVI.

CONCERNING THE CITY OF SAPURGAN.