A third skull of very similar character, except that it is not so much depressed, has come under my observation. It was discovered in a submarine or, rather, subterranean peat-bed or ancient forest, 30 feet below the sea-level, at Sennen, near the Land’s End, in Cornwall; and a brief notice and outline figure of it will be found in the “Natural History Review” for 1861.[114] The Sennen skull has the same elongated form; but it is higher than either the Cefn, St. Asaph, or Borris crania, having an altitudinal index of ·730. On the whole, these three skulls (_i.e._ those from Borris, Sennen, and St. Asaph) would appear to have a common character, and to be of a different type from either the Perthi-Chwareu or the Mewslade form. As a rule it may, I think, be stated that in all brachy-cephalic skulls the breadth exceeds the height, whilst the reverse is the case in the dolicho-cephalic. Individual exceptions are of course not unfrequently met with, more especially among very mixed races, such as the modern English; but I am myself acquainted with only two dolicho-cephalic _races_, properly so termed, in which the rule does not hold good. These are the Tasmanian (not Australian) and the Bushman. Any exceptions, therefore, to either rule among ancient and, consequently, less mixed races are worthy of being noted. As regards modern brachy-cephalic skulls the law holds almost universally, the only marked exception, except in an individual here and there, being in two Karén skulls, in which, although both decidedly brachy-cephalic, the respective indices stand as ·848 to ·924, and as ·790 to ·842. Among priscan brachy-cephalic skulls the most remarkable and important exceptions I have met with occur among the neolithic crania in the Copenhagen Museum, more than half of which are brachy-cephalic, and most of the others nearly so, the mean cephalic index of 21 skulls being ·790, whilst the mean altitudinal is as high as ·810. In fact, out of 12 skulls whose indices vary from ·795 to ·838, no fewer than 10 have the latitudinal index less than the altitudinal. The exceptions to the rule as applied to dolicho-cephalic skulls also appear to be far more common among the ancient than among the modern, excepting the two races I have above referred to. In a long list of ancient and priscan skulls, I find the following having the tapino-cephalic character:--