No. 2 is a piece of cream-colored limestone, with numerous minute, scattered, rhombic, striated, ganoid scales, which I cannot venture to name, though I believe them to be palæoniscid. Associated with these is a small portion of the margin of a jaw, with numerous minute sharp conical teeth. But also lying among these minuter relics is a scale of a much larger size, and clearly belonging to another fish. It measures 1-4 inch in height by the same in breadth; its shape is rhomboidal, having an extensive anterior covered area, and a strong articular spine projecting from the upper margin. The free surface is brilliantly ganoid, and marked with furrows separating feeble ridges, which pass rather obliquely downwards and backwards across the scale, and terminate in eight sharp denticulations of the hinder margin. A little way off is the impression of the attached surface of a similar scale, and there are also two interspinous bones, probably belonging to the same fish.
This is probably also a palæoniscid scale, resembling in shape those of Acrolepis, but it is rather thinner than is usually the case in this genus. It has also considerable resemblance to some of those scales from the European Trias, named by Agassiz Gyrolepis. Though it may be rather venturesome to name a species from such slender material, nevertheless we may, provisionally at least, recognize the scale as Acrolepis (?) Africanus.
Nos. 3 and 4 are small pieces of the same limestone, covered with the minute striated palæoniscid scales referred to above.
No. 5 is a piece of gray micaceous shale, with scales of yet a fourth species of palæoniscid fish. One conspicuous scale unfortunately, like all the rest, seen only from the attached surface, is 1-4 inch in height by nearly 1-6 in breadth; it is tolerably rectangular in shape, having a well-developed articular spine and fossette. Part of the scale is broken away at the anterior margin, the impression brought into view showing that the covered area is narrow, and indicating that the free surface is striated with rather sharp ridges passing obliquely across the scale. The posterior margin is finely denticulated.
Though this scale is in my opinion specifically, and possibly generically, distinct from those previously named, the outer surface not being properly displayed renders it impossible to give a sufficient diagnosis.
No. 6 is a piece of the same shale, having the clavicle of a small palaeoniscid fish, which it is, however, impossible to name.—I am, yours faithfully,
R. H. TRAQUAIR.
These fossiliferous beds seem to occupy a comparatively limited area, and have a very high dip in a south-easterly direction. At the spot where my observations were taken they did not extend over more than half a mile of country, but it is possible that the formation may persist for a long distance in other directions. Indeed, I traced it for some miles in the direction in which, some fifty or sixty miles off, lay the coal already described, and to which it may possibly be related.
With one or two general remarks upon surface geology and physical geography I bring this note to a close. First, regarding the Lakes Nyassa and Shirwa,—there is distinct evidence, and especially in the case of the latter, that they have formerly occupied a considerably larger area than at present. Shirwa is an extremely shallow lake; though the eastern and southern shores are mountainous, it is suggestive rather of an immense bog than of a deep inland sea. For many miles before reaching the shore there are signs that one is traversing the site of a former and larger Shirwa, which may possibly at one time have been actually connected with the lower extremity of Lake Nyassa. To substantiate this conclusion, however, will require more detailed examination of the Shiré Highlands than I was able to give. The peculiarity of Shirwa is that the water is brackish to the taste, while that of Nyassa and of the other Central African lakes, with the exception of Lake Leopold, is fresh. The shallowness of Shirwa, and the precariousness of its outlet through Lake Cheuta to the Lujenda, amply account for this difference; for the narrow waters of Nyassa and Tanganyika are thoroughly drained and profoundly deep.
That Lake Nyassa is also slowly drying up is evident from the most superficial examination of its southern end. There it has already left behind a smaller lake—Lake Pomalombé—a considerable expanse of water, through which the Shiré passes a few miles after emerging from Lake Nyassa, but already so shallow that nowhere in the dry season does the depth exceed three fathoms. If the silting up of this lake continues for a few years it will render this sheet of water, which commands the entrance to Lake Nyassa, totally unnavigable, and thus close the magnificent water-highway at present open, with a portage of seventy miles, from the top of Lake Nyassa to the Indian Ocean at the mouth of the Zambesi.