February 27, 1891.] 



SCIENCE. 



121 



as to throw their shadows on a new portion of the rod-and-cone 

 layer. From the nature of the case, the corpuscles cannot be 

 rendered invisible, like the capillaries. 



The phenomena described above were first observed by the 

 writer a dozen years ago; and, though it is probable that others 

 have observed the same, consultation with persons and books 

 that would be likely to furnish the information of such knowledge 

 have shown that these facts are either unknown, or at least not 

 generally known. That the facts here published may be observed 

 by any one seems proved by the fact that they have been corrob- 

 orated by almost every one who has made the attempt under the 

 writer's direction. J. E. Todd. 



Tabor College, Talior, Ic, Feb. 16. 



Classification of American Languages. 



In your issue of Feb. 6 appears an article by Major J. W. Pow- 

 ell, chief of the Bureau of Ethnology of the Smithsonian Institu- 

 tion, on the study of what he calls " Indian " languages, with a 

 list of families in the United States. 



This article contains statements so much at variance with the 

 leading authorities in linguistic science, that they should not be 

 allowed to pass in silence. 



In the first place, the term " Indian languages," applied to those 

 spoken by the native tribes of this continent, is a misnomer based 

 on an ancient blunder, and lias been repudiated by all modern 

 writers of weight. The so-called "Indians" are the "American 

 race," and their languages are "American languages," by the 

 common consent of ethnographers. Is the Bureau of Ethnology 

 a sanctuary for the preservation of exploded errors, that it throws 

 its influence into the scale to perpetuate this discarded blunder? 



Much of the article alluded to is devoted to explaining and de- 

 fending the nomenclature adopted by the bureau. In several 

 points it requires still further defence. The arbitrary assumption 

 of the date 1336, anterior to which the " law of priority " is 

 decreed not to hold good, is not justified by the reasons given. 



The dictum that "no family name shall be recognized if com- 

 posed of more than one word," is not merely arbitrary, but has 

 nothing in its favor and much against it. Frequently a class- 

 name compounded of two words is particularly useful, as convey- 

 ing a much wider idea than a single word. This is f uUy recog- 

 nized by the best linguists of the day. Thus, Friedrich Miiller 

 employs the terms " Indo-Germanic," "Ural-Altaic," etc. The 

 reasons assigned for rejecting such compounds are quite inade- 

 quate, and contrary to the practice of the highest authorities. 



The adoption of the termination an or ian to denote families or 

 stocks of languages is not original with our Bureau of Ethnology, 

 though the article miglit lead the reader to suppose it a new de- 

 vice. Some writers adopted it long before tlie bureau was organ- 

 ized, but the plan did not meet with general approval. The 

 cacophony of such words as " Eskimauan," "Muskhogean," etc., 

 in Major Powell's list is apparent to every one who has not had 

 the advantage of that training by the bureau to which he refers 

 with pride as destroying all sense of euphony. 



But the portion of the article in question which will most com- 

 pletely "knock the wind" out of those old-fogy linguists in 

 Europe, and those in our own country who have been reared on 

 Aryan and Semitic tongues, is Major Powell's declaration that 

 " grammatic similarities are not supposed to furnish evidence of 

 cognation ; " that in his classification grammatic structure has 

 been neglected, and lexical elements only considered. 



Now, if it were said that in most instances we are obliged to 

 depend on lexical elements because the grammatic structure has 

 not been ascertained, the position would be sound and in accord 

 with the recognized principles of the science of language ; but to 

 place the words of a tongue above its grammar in instituting 

 comparisons is a feat of sucli daring or of such ignorance, that it 

 requires a man long accustomed to frontier life to venture it. If 

 there is any one principle in modern linguistics which we may 

 look upon as thorouglily established, it is that the grammatic 

 framework of a language is incomparably more stable than its 

 lexicon. If there has ever been an instance where a language of 

 agglutination has changed into one of inflection, it is not recorded 



" in the books." It is precisely the grammar which is the perma- 

 nent part of a language, and not its vocabulary. Modern Turkish 

 has borrowed three-fourths of its words from Arabic, Greek, Per- 

 sian, etc. ; but its grammar remains almost precisely that of the 

 pure stock, the Yakut of the delta of the Lena. This principle is 

 as true of American tongues as of others, and the evidence of it 

 has been abundantly set forth by Friedrich Mtiller and Lucien 

 Adam. D. G. Beinton, M.D. 



Philadelphia, Penn., Feb. 20. 



The Food of Moles. 



It is stated in the " Encyclopaedia Britannica" that moles are 

 entirely carnivorous, are exceedingly rapacious, and will die if 

 left longer than eight or ten hours without food. I recently kept 

 a living mole for a time to study its habits. I shut it in a venti- 

 lated wooden box, giving it a tin lid full of water, and some 

 grains of corn. It drank the water, refused the corn, and, while 

 kept strictly in the dark, was quiet. After twelve hours' captivity 

 I offered it boiled rice, which it refused. After sixteen hours' 

 fasting, it ate bread and milk, though not freely. When I had 

 had it twenty hours, I gave it cracked oats, soaked well in milk, 

 but uncooked. This it ate ravenously. I then released it in the 

 room, and it travelled about, seeking a place to burrow, and made 

 itself troublesome tearing at the carpet and upholstery. I threw 

 down a large thick woollen mitten, which it speedily found and 

 entered, tlirusting its head into the thumb. If undisturbed, it 

 would hide in this way for hours, the light and warmth of the 

 room seeming greatly to annoy it. It lived in the mitten for 

 three days, coming out to eat oats soaked in milk, but refusing 

 cooked oats. It was given one small meal of raw meat. At the 

 end of four days it was killed, being apparently in a healthy con- 

 dition, and not having lost any flesh. 



Julia McNaie Wright. 



Fulton, Mo., Feb. 30. 



Cold and Warm Waves. 



Two rival theories have been propounded recently regarding 

 the origin of the waves or masses of cold air which appear to 

 traverse the country toward the east. One of these finds the 

 source of cold in the upper regions of the atmosphere, and consid- 

 ers that the cold air above mixes with that below, and thus 

 gradually approaches the earth's surface. Those supporting the 

 other theory, however, deny that any considerable cold can be 

 brought down in this way, because the compression to which the 

 air would be subjected would heat it, but they claim that the cold 

 is due to the radiation of heat through the very clear sky which 

 is a well-nigh invariable accompaniment. Without expecting to 

 establish the exact truth in this matter, it has yet seemed a sub- 

 ject of much importance ; and it may be well, at this stage in the 

 discussion, to set forth a few facts that may be of use in the final 

 solution of the problem. 



Those who have been making forecasts of the weather have 

 recognized for more than a dozen years three great classes of 

 temperature falls: 1. Those which come with the advance of areas 

 of high pressure; 3. Those which follow immediately in the rear 

 of great storms independently of any high area ; 3. Those which 

 occur under a combination of these two causes. It should be 

 noted that the first two classes do not invariably occur even when 

 the conditions seem favorable, and great care is needed in exam- 

 ining other conditions, which, though apparently remote, may 

 yet become exceedingly important factors in the development of 

 the cold wave. The occurrence of the cold is independent of the 

 wind, though the extent of the wave is markedly dependent on 

 the rapidity of its advance, and a rapid motion has a tendency to 

 increase the wind. Some have thought that the wind brings the 

 cold ; but this cannot be the case, for often there is no wind, or at 

 least it rarely attains fifteen miles per hour, while the cold wave 

 advances at double that velocity. One of the essential conditions 

 needed for a cold wave is an elimination of the moisture in the 

 air, and this removal of moisture is oftentimes very remarkable. 

 In one case three-fourths of this moisture was removed in 110 



