When we compare the highest and lowest races with each other, the contrast becomes still more striking, viz: the Teutonic with the Hottentot and Australian. The former family gives a mean capacity of 92 inches, while the latter two yield but 75, or a difference of _17 cubic inches_ between the skulls of these types! Now, as far back as history and monuments carry us, as well as crania and other testimonies, these various types have been _permanent_; and most of them we can trace back several thousand years. If such permanence of type through thousands of years, and in defiance of all climatic influences, does not establish _specific_ characters, then is the naturalist at sea without a compass to guide him. These facts determine clearly the arbitrary nature of all classifications heretofore adopted; the Teuton, the Jew, the Hindoo, the Egyptian, &c., have all been included under the term _Caucasian_; and yet they have, as far as we know, been through all time as distinct in physical and moral characters from each other, as they have from the negro races of Africa and Oceanica. The same diversity of types is found among all the other groups, or arbitrary divisions of the human family. Rich and rare as is the collection of Dr. Morton, it is very defective in many of its divisions, and it occurred to me that this deficiency might to some degree be supplied by the hat manufacturers of various nations; notwithstanding that the information derived from this source could give but one measurement, viz: the _horizontal periphery_. Yet this one measurement alone, on an extended scale, would go far towards determining the general size of the brain. I accordingly applied to three hat dealers in Mobile, and a large manufacturer in New Jersey, for statements of the relative number of hats of each size sold to adult males; their tables agree so perfectly as to leave no doubt as to the circumference of the heads of the white population of the United States. The three houses together dispose of about 15,000 hats annually. The following table was obligingly sent me by Messrs. Vail & Yates, of Newark; and they accompanied it with the remark, that their hats were sent principally to our Western States, where there is a large proportion of German population; also that the sizes of these hats were a little larger (about one fourth of an inch) than those sold in the Southern States. This remark was confirmed by the three dealers in Mobile. Our table gives, 1st. The number or size of the hat. 2d. The circumference of the head corresponding. 3d. The circumference of the hat; and lastly, the relative proportion of each No. sold out of 12 hats. Size--inches. Circum. Circum. Relative of head. of hat. prop. in 12.