200 THE TARIM RIVER.
have nothing to contend against except sand and silt. For the ground has been already »worked» by the older loop; that is to say, it is already ploughed up and loosened, and there is no vegetation left. Hence the situation shown in D cannot be of any long duration; indeed the change shown in E will be brought about in the same high-water season that witnesses the breach. And as a matter of fact we do find, in the case of the great majority of these boldschemals, that the river has worn away to a gently curved line (a in fig. E) the sharp extremity of the island inclosed by the boldschemal. But whilst the boldschemal itself has been undergoing metamorphosis the island which forms its nucleus has remained unchanged, and the resistance now offered by its vegetation compels the water that is left in the boldschemal to take the shape depicted in fig. F. As time goes on, the water area of the boldschemal steadily decreases; this stadium is exhibited in, for instance, the boldschemal shown in fig. 165 and 166. In nearly every case there is an interval between the two blunted horns of the river which penetrate the upper and lower extremities of the boldschemal respectively; it is very rarely indeed that they both coalesce into one, as in fig. D.
Let us follow the development of the loop yet one stage more from fig. F. In fig. G we have the newly formed loop cut off in its turn, and now forming a boldschemal. Fig. 167 shows a double boldschemal of this kind, which we passed at Ja-kotan on the 2oth October, and fig. 168 another below Dängsur-köl, Dec. 6th. This will show that I have not deduced this later stage as a result of reasoning, but that the condition actually does exist, although I admit instances of this particular formation are extremely few. The explanation is that, in the majority of cases, when boldschemal No. 2 is ripe for being cut off, boldschemal No. 1 is already dried up, and choked with sand and vegetation.
Upon comparing the first and the last fig. in our series, it is almost impossible to recognise in G. the self-same loop that is shown in fig. A; nevertheless it has gone through all the intermediate stages of metamorphosis. Yet G is just as far from being the final stage as A was. In fact, there exists only one reason why the process should not go on for ever, and that is the fickleness of the river. The only portions of the river's course in which one can imagine loops exhibiting yet
IIIF Color HighRes
