[The Rev. C. Holcombe writes (pp. 64-65):  “P’u-chau fu lies on a level with the Yellow River, and on the edge of a large extent of worthless marsh land, full of pools of brackish, and in some places, positively salt water....  The great road does not pass into the town, having succeeded in maintaining its position on the high ground from which the town has backslided....  The great road keeping to the bluff, runs on, turning first south, and then a trifle to the east of south, until the road, the bluff, and Shan-si, all end together, making a sudden plunge down a precipice and being lost in the dirty waters of the Yellow River.”—­H.C.]

Not maintaining the infallibility of our traveller’s memory, we may conceive confusion here, between the recollections of his journey westward and those of his return; but this does not remove all the difficulties.

The most notable fortress of the Kin sovereigns was that of T’ungkwan, on the right bank of the river, 25 miles below P’u-chau fu, and closing the passage between the river and the mountains, just where the boundaries of Ho-nan, Shan-si, and Shen-si meet.  It was constantly the turning-point of the Mongol campaigns against that Dynasty, and held a prominent place in the dying instructions of Chinghiz for the prosecution of the conquest of Cathay.  This fortress must have continued famous to Polo’s time—­indeed it continues so still, the strategic position being one which nothing short of a geological catastrophe could impair,—­but I see no way of reconciling its position with his narrative.

[Illustration:  Plan of Ki-chau, after Duhalde.]

The name in Ramusio’s form might be merely that of the Dynasty, viz. Tai-Kin= Great Golden.  But we have seen that Thaigin is not the only reading.  That of the MSS. seems to point rather to some name like Kaichau.  A hypothesis which has seemed to me to call for least correction in the text is that the castle was at the Ki-chau of the maps, nearly due west of P’ing-yang fu, and just about 20 miles from the Hwang Ho; that the river was crossed in that vicinity, and that the traveller then descended the valley to opposite P’u-chau fu, or possibly embarked and descended the river itself to that point.  This last hypothesis would mitigate the apparent disproportion in the times assigned to the different parts of the journey, and would, I think, clear the text of error.  But it is only a hypothesis.  There is near Kichau one of the easiest crossing places of the River, insomuch that since the Shen-si troubles a large garrison has been kept up at Ki-chau to watch it.[1] And this is the only direction in which two days’ march, at Polo’s rate, would bring him within 20 miles of the Yellow River.  Whether there is any historic castle at Ki-chau I know not; the plan of that place in Duhalde, however, has the aspect of a strong position.  Baron v.  Richthofen is unable to accept this suggestion, and has favoured me with some valuable remarks on this difficult passage, which I slightly abridge:—­