394 



SCIENCE. 



[Vol. XIV. No. 357 



containing about two hundred octavo pages, will be issued each 

 year. " The contributors will be, for the most part, instructors in 

 the university, or graduates of the same, but contributions of other 

 scholars will not be absolutely excluded. Any correspondence 

 respecting contributions should be addressed to Professor James 

 B. Greenough, Professor John Williams White, or Professor F. D. 

 Allen, Cambridge, Mass. Subscriptions (one dollar, four marks, 

 or five francs a 'volume) may be sent to Otto Harrassovvitz 

 Leipzig, Germany ; Ginn & Co., 57 & 59 Ludgate Hill, London 

 E.C., England ; or the latter firm at Boston, New York, or Chicago 

 — With the December number the Magazine of American His- 

 ioty completes its twenty-second volume. The frontispiece to the 

 current issue is a -portrait of Lord Brougham; and the opening 

 paper by the editor is a sketch of his early career during the in- 

 fancy of our Republic, with pen-pictures of his contemporaries and 

 surroundings, the establishment of the Edinbtcrgh Review, and 

 the marriage of its editor in New York City. The second illus- 

 trated paper is a "Tribute to Hooper C. Van Voorst," the late 

 president of the Holland Society, by George W. Van Siclen. The 

 third contribution is " The Story ol Brave, Beautiful Margaret 

 Schuyler," an historic ballad from the pen of Judge Charles C 

 Nott of Washington. Curiously interesting is the article following, 

 of R. W. Shufeldt, " The Drawings of a Navajo Artist," illustrated 

 with the Indian pencil ; as is also the " Acrostic by John Ouincy 

 Adams," in facsimile, from Ella M. M. Nave. "The Sciota Pur- 

 chase in 1787," by Col. E. C. Dawes of Cincinnati, and the "Private 

 Contract Provision in Ordinance of 1787," by Hon. W. P. Cutler, 

 are important contributions to the number. These are ably 

 written, and will doubtless serve to correct many errors in recent 

 histories of Ohio. " Joseph Hawley, the Northampton States- 

 man," is the theme of a paper by Charles Lyman Shaw; " Fort 

 Perrot, Wisconsin," is from T. H. Kirk; "First Editions of the 

 Bible printed in America," from Clement Furgeson ; and "Gen. 

 Grant and the French," from Theodore Stanton of Paris. This 

 magazine is steadily exerting an educational and healthful influence 

 in all departments of literature and study. 



was repeatedly tried with the same result. Their movements on 

 being disturbed very forcibly reminded me of a child when sud- 

 denly waked out of sound sleep." JAS. Lewis Howe. 



Louisville, Ky., Nov. 2i. 



LETTERS TO THE EDITOR. 



^^^Correspondents arereguestedtoba as brief as possible. The writers name is 

 in all cases reqaired as proof of eood faith. 



The editor mill be glad to publish any queries consonant with the character of 



On request, twenty copies of the member containing his communication will be 

 furnished free to any correspondent. 



Intelligence of Ants. 



I SEND you the following regarding ants, by Mr. W. E. Bos- 

 worth of this city, written out at my request, which seems to me 

 an interesting and at the same time somewhat rare observation. It 

 is almost exactly similar to the account by McCook of the sleeping 

 of harvesting ants, of Texas, as quoted in G. J. Romanes' " Animal 

 Intelligence," p. 84. I do not recall any other instance given of 

 the .sleeping of ants. " At different times, and for more than one 

 season, I was favorably situated to see the movements of quite a 

 large colony of small black ants, as they passed to and fro in their 

 busy haste over a board floor, going, as I supposed, for their sup- 

 ply of water, which was in the direction of a small stream close by. 

 While watching their quick, eager movements, there were several 

 along the line that attracted my attention, as they remained in one 

 place so long that I concluded they must be dead ; and although 

 they were directly on. the line of march, and in the way of the 

 others, these passed on, paying no attention to them whatever. 

 At another time I noticed that one of the ants supposed to be dead 

 got up, and walked off as lively as the rest ; and, while watching 

 this one, another one close by began to slow up, seemed to totter 

 in his gait, and finally came to a dead halt. After seeing this, it 

 occurred to riie that the one had just waked up, and the other had 

 just^gone to sleep. In order to test the matter, and gratify my 

 curiosity, I concluded to experiment on some of them. With a 

 fine straw they were gently rubbed on the back. This mild treat- 

 ment did not make the slightest impression on them ; but a sharp 

 push seemed to take them completely by surprise, and to fully 

 arouse them. For an instant they seemed lost, circulating around, 

 running up and down, but finally starting off with the rest. This 



Galton's Bodily Efficiency Diagram and the Marking System.. 

 Francis Galton's bodily efficiency diagrams {NaUtt-e, Oct. 

 31, i88g) can evidently be applied to the rating, on an arbitrary- 

 scale, of all sorts of things besides physical measurements and tests. 

 For instance ; the annexed diagram represents, by Galton's- 

 method, the rating of errors as the measure of precision gradually 

 rises. The data were taken from the table on p. 12 of Merriman's- 

 " Least Squares " (first edition). The curves are drawn in general 

 for values of .r differing by .1 ; the ordinates in all cases being 

 values of h, and the abscissas the rating on a scale of 100. The 

 diagram shows at a glance how in all cases the rating of the same 



error decreases as the measure of precision increases, but how, for 

 very large and very small errors (see the curves x = .01 and 

 X = 1.7), the measure of precision affects the rating little. 



The rating of any errors which are distributed roughly according 

 to the probability curve, as they are, for instance, in every school 

 examination, ought to conform in general to these curves, and I 

 think teachers usually strive to have it do so, either consciously or 

 instinctively. If the error is flagrant, the question containing it is 

 marked zero, or nearly so. The discrepancies in the marks of dif- 

 ferent teachers, or in the marks of the same teacher at different 

 times, seem due to the different measures of precision mentally 

 adopted. The curves show that these variations of the measure 

 of precision affect most the rating of mediocre work, and this also 

 accords witb the experience of teachers. Now, of course the errors 

 of each scholar have their own probability curve and their own 

 value of h, which perhaps might be calculated from a long series 

 of examination-papers. It would probably differ for different sub- 

 jects. The custom, then, of marking good and poor scholars on 

 different scales has a reason. The only question is, whether these 

 scales can be so systematized as to be quite just, and whether it 



