sand, the southern bay of the Tuvadake-köl would ' form the first bajir of the Begelik-köl.
Taken as a whole the Begelik-köl is 11.6 km. long, with a maximum breadth of 4.15 km. It is consequently appreciably larger than the lakes I have described above; and it likewise differs from them in its peculiar and fantastic shape. For whereas they are more homogeneous and self-contained, and either elliptical in shape or composed of two basins, or, again, resemble a section of a river-bed, the Begelikköl is, as it were, plucked asunder, and it is possible that the two bays, the southern bay of Tais-köl and the Ku-tuttu, are both more irregular and penetrate deeper inland than I was able to perceive from the canoe. Otherwise all the characters which we have found to distinguish the desert lakes recur also here. The separate members of the lake are prolonged, if not in the usual north-north-east to south-south-west direction, at all events from north to south. The Ku-tuttu and Avul Nias Bekne-kakmasi are even prolonged from north-west to south-east, and this direction, taking the prevailing winds into account, is a more reasonable one than the usual north-north-east and south-south-west.
Another feature common to the Begelik-köl and its associates is that it too has its greatest depth on its eastern side, under the highest dunes, while it shallows towards its western shore; this is especially true of the Tais-köl. The extreme north of the Aghis-köl is likewise shallow. In this part of the lake the isobathic lines for 2 and 4 m. run a long way out from the shore, whereas on the east side of the lake they come close together under the dunes. The shape of the lake is of course determined by the configuration of the underlying substratum and the arrangement of the dunes. To judge from the contours of the Begelik-köl neither the one nor the other of these determinants is so precise or so distinct as in the case of the other lakes, where it is almost everywhere possible to distinguish two long and stupendous accumulations of dunes, one on each side of the lake. Here too there are indeed accumulations of sand, but they are not so accurately arranged, and are broken and truncated. Even with the naked eye it is easy to perceive that these dunes are not appreciably lower than the dunes beside the lakes we have already considered. Neither is their parallelism so regular: on the east side of the Tuvadake-köl there is a dune running north and south, and another on the west of Nias Bek's bay and the Tais-köl, although in the north it inclines slightly to the north-north-west. On the peninsula east of the Ghol-köl there is a dune which lies distinctly south-southeast and north-north-west. Thus the Begelik-köl lies in a part of the desert in which the sand is not only lower but less regularly cohesive than in other parts; and it is to this circumstance that we must attribute the fact of the greater depth of these lakes. East of the easternmost part of the Tuvadake-köl the conformation of the sand allows comparatively free access to the wind, and there can be no doubt that the particular locality was formerly, at a time when there was no water in the lake-basin, exposed to more energetic wind erosion. As I shall have to discuss the sandy desert and its extent in another part of this present work, I will only say here, that the increasing depth in the lakes as we descend the river is not merely a matter of chance, due to the small number of soundings taken, but it is due to other causes. The maximal depths of the four lakes I measured are as follows: —
IIIF Color HighRes
