io8 



SCIENCE 



[Vol. XVIII. No. 446 



eral reading. It is also frequently stated that the vocabulary of 

 certain miners consisted of but one hundred words. Whether 

 this was an actual count or merely an estimate I do not know, but 

 should think that it must be the latter. In order to determine the 

 size of an ordinary vocabulary I could think of no better means 

 than to find out the number of words used in some standard work 

 that is easily read and understood by everybody. Nor could I 

 think of any book better suited for the purpose in view than that 

 great English classic, " Robinson Crusoe." The copy of that work 

 in my possession contains 460 pages, and I first noted down all of 

 the different words found on every tenth page (counting as a sep- 

 arate word what is given as such in the dictionary). This proba- 

 bly gave more different words than forty-six consecutive pages 

 would have done, because a greater number of subjects and inci- 

 dents are discussed and described. I then noted the new words 

 on the remaining nine of a section of ten pages in the front part 

 of the book, and then of a section in the latter part, in order to 

 get a basis for estimating the new words in the rest of the book. 

 The number of words on the sixty-four pages counted was thirty- 

 one hundred, and if the percentage of decrease for each section of 

 nine pages from the section counted just before it should be the 

 same as for the two sections counted, there would be on the re- 

 maining 396 pages about three thousand words. It may be, as 

 would seem probable, that the percentage of decrease would in- 

 crease after awhile, but so far as counted there was no sign of an 

 increased rate of falling off. The falling off was very rapid for 

 the first five pages, less rapid for the nest twenty, and after that 

 not enough to be evident unless the average of a number of pages 

 was taken. It seems quite certain, then, that De Foe, in writing 

 his account of the adventures of Robinson Crusoe, used not less 

 than five or six thousand words. Children of ten or twelve years 

 read the book with pleasure, and probably have a pretty clear 

 idea of the meaning of nine out of ten of the words they find in 

 it. The work probably contains most of the verbs and a large 

 proportion of the adjectives and adverbs in common use, but there 

 is a large number of nouns, both common and proper, familiar to 

 every child, which De Foe had no occasion to use in this work. 

 It is probable then that to read ordinary general reading in English 

 understandingly one needs to be familiar with from six to ten 

 thousand words. The same must be true for other languages 

 equally rich in synonymes. Grimm's " Marchen " contains a 

 vocabulary of between four and five thousand words, yet any one 

 who can readily read those stories needs a dictionary constantly 

 by his side when reading ordinai-y German. 



From the data at hand I should estimate the vocabulary of a 

 citizen of the United States with a common-school education and 

 of ordinary intelligence and reading at about ten thousand words, 

 and that of a well-read college graduate, and of those who have 

 pursued a university course, at from twenty thousand upwards to 

 perhaps one hundred thousand. One's vocabulary is usually 

 nearly complete at thirty years of age. If but two words are 

 learned each day the vocabulary at that age would be only twenty 

 thousand. My records show that young children acquire new 

 words more rapidly than that. 



As to the composition of a vocabulary, I find that in the dic- 

 tionary about 60 per cent of the words are nouns, a little over 22 

 per cent adjectives, and a half that per cent verbs, and a fourth 

 adverbs. Pronouns, prepositions, and conjunctions, though used 

 in every sentence, constitute a very small part of a general vocabu- 

 lary — none were found in examining fifteen pages, or one in every 

 hundred, in the dictionary. Of the thirty-one hundred words 

 obtained from " Robinson Crusoe," a little over 45 per cent were 

 nouns, 24 per cent verbs, a little over 17 per cent adjectives, and 

 7 per cent adverbs. Probably nearly every one is familiar with a 

 larger proportion of the verbs than of the nouns in the dictionary, 

 but "Robinson Crusoe" is particularly rich in verbs. Many of 

 them are used only as participles, the form in many cases being 

 the same as for the adjectives, but they only counted as verbs un- 

 less distinctively used as adjectives. As already suggested, the 

 ordinary vocabulary contains a larger proportion of nouns than 

 are found in "Robinson Crusoe," and many that are not found 

 in the dictionary, although the proportion is probably not greatly 

 different from what it is in the latter. In small vocabularies the 



proportion for the different parts of speech is quite different. Of 

 the 215 words on the first page of "Crusoe" that I counted, 5 

 per cent were prepositions, 10 per cent adverbs, 10 per cent pro- 

 nouns, 6 per cent conjunctions, and but 34 per cent nouns. This 

 must be borne in mind in considering small vocabularies like those 

 of children. 



As a matter of some general interest, and a point of considerable 

 importance, in considering the question of the pronouncing vocabu- 

 lary of children, it is worth while to notice with what letters of 

 the alphabet the greatest number of words begins. The letters s, 

 p, and c begin nea'rly one-third of the words in the English lan- 

 guage. The following is the order for the letters most frequently 

 used in the dictionary : s, p, c, a, t, b, r, m, d, f, e, h, 1, g, w, o, 

 V, n, u; in "Robinson Crusoe," s, c, p, a, f, b, r, m, e, t, w, h, 1, 

 i, g, o, n, u, V. 



Further data are needed in order to confirm or correct the esti- 

 mates given in this article. E. A. KmKPATRlCK. 



Rhodes, Iowa, Aug. 14. 



Climatic Changes in the Southern Hemisphere. 



Having had occasion to cruise a considerable time over the 

 Southern Ocean, I have had my attention directed to its prevail- 

 ing winds and currents, and the way in which they affect its 

 temperature, and also to the ice-worn appearance of its isolated 

 lands. 



It is now generally conceded that the lands situated in the high 

 latitudes of the southern hemisphere have in the, remote past been 

 covered with ice sheets, similar to the lands which lie within the 

 antarctic circle. The shores of southern Chili, from latitude 40' 

 to Cape Horn, show convincing evidence of having been overrun 

 by heavy glaciers, which scoured out the numerous deep channels 

 that separate the Patagonian coast from its islands. The Falkland 

 Islands and South Georgia abound with deep friths ; New Zealand 

 and Kerguelen Land also exhibit the same evidence of having 

 been ice-laden regions; and it is said that the southern lands of 

 Africa and Australia show that ice accumulated at one time to a 

 considerable extent on their shores. At this date we find the 

 southern ice-sheets mostly confined to regions within the antarctic 

 circle; still the lands of Chili, South Georgia, and New Zealand 

 possess glaciers reaching the low lands, which are probably grow- 

 ing in bulk ; for it appears that the antarctic cold is slowly on the 

 increase, and the reasons for its increase are the same as the causes 

 which brought about the frigid period which overran with ice all 

 lands situated in the high southern latitudes. 



Why there should be a slow increase of cold on this portion of 

 the globe is because of the independent circulation of the waters 

 of the Southern Ocean. The strong westerly winds of the southern 

 latitudes are constantly blowing the surface waters of the sea from 

 west to east around the globe. This causes an effectual barrier, 

 which the warm tropical currents cannot penetrate to any great 

 extent. For instance, the tropical waters of the high ocean levels, 

 which lie abreast Brazil in the Atlantic and the east coast of Africa 

 in the Indian Ocean, are not attracted far into the southern sea, 

 because the surface waters of the latter sea are blown by the 

 westerly winds from west to east around the globe. Consequently 

 the tropical waters moving southward are turned away by the 

 prevailing winds and currents from entering the Southern Ocean. 

 Thus the ice is accumulating on its lands, and the temperature of 

 its waters slowly falling through their contact with the increasing 

 ice; and such conditions will continue until the lands of the high 

 southern latitudes are again covered with glaciers, and a southern 

 ice period perfected. But while this gathering of ice is being 

 brought about, the antarctic continent, now nearly covered with 

 an ice-sheet, will, through the extension of glaciers out into its 

 shallow waters, cover a larger area than now ; for where the waters 

 are shoal the growing glaciers, resting on a firm bottom, will ad- 

 vance into the sea, and this advancement will continue wherever 

 the shallow waters extend. Especially will this be the case where 

 the snowfall is great. 



Under such conditions, it appears that the only extensive body 

 of shallow water extending from the ice-clad southern continent 



