204 



SCIENCE. 



[Vol. XIII. No. 319 



— Houghton, Mifflin, & Co. have just issued a new Hfe-size por- 

 trait of Dr. Holmes, which is even better than the earher one. 



— Charles Scribner's Sons have in preparation a handsome pop- 

 ular Hbrary edition, in four i2mo volumes, of Bourrienne's well- 

 known " Memoirs of Napoleon," a standard work of which many 

 hundreds of imported sets have beeii sold every year. This new 

 ■edition will be an exact reprint of the latest English edition, and 

 will contain the thirty-eight portraits of the original, together with 

 all the other features that give distinction to the work. The price 

 "will be sufficiently low to bring these volumes within reach of all 

 would-be readers. 



— G. P. Putnam's Sons have in press, as their own commemora- 

 tion of the centennial anniversary of the inauguration of Washing- 

 ton, a unique limited edition of Irving's " Life of Washington," — a 

 work for which Bryant predicted " a deathless renown." The set 

 "will be issued in five volumes, handsomely printed in large quarto 

 form, and will contain 200 illustrations, comprising 130 steel 

 plates and 70 woodcuts printed on India paper and inlaid in the 

 text. The plates include portraits of all the noteworthy generals 

 -and statesmen of the American Revolution. But 300 sets will be 

 -issued, and the type will be distributed as printed from. The price 

 to subscribers has been fixed at $50. 



— Mrs. Stowe has been able to revise the biography of herself, 

 written by the Rev. Charles Stowe and Mr. Kirk Munroe. It will 

 ■be published at an early day by Houghton, Mifflin, & Co. 



— Baron Grancy will shortly issue, in Paris, a volume on Ameri- 

 can customs. It is to be in the shape of a novel, to be entitled " A 

 Trench Ranch in Dakota," and will treat wholly of Dakotan affairs. 

 The author, according to a despatch to the New York IVorld, is 

 the original founder of the Fleur de Lys settlement of French 

 ■horse-breeders, whose life in Dakota this book is meant to de- 

 scribe. 



LETTERS TO THE EDITOR. 



'*tCorrestonde>itsarereqtiestedtobeasbriefasjiossible. The writer's name is 

 in allcases regtiired as proof of j;ood faith. 



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

 4he journal. 



Twenty copies of the number containing his communication tuill be furnished 

 'free to any correspodent on request , 



An Acoustic Mill. 



When a vibrating tuning-fork is brought near to a light body, 

 like a pith-ball or a small piece of paper, the latter moves towards 

 the fork as if attracted by it. This phenomenon was observed by 

 Guyot in 1832, and was rediscovered by Guthrie in 1870. It has 

 been supposed by some that gravitation could be explained by the 

 vibratory motions, such as atoms and molecules are known to have ; 

 "but it does not appear that gravitation sustains any quantitative 

 relation whatever to the temperature of a body, such as would be 

 the case if molecular vibration was the cause of it. The observed 

 phenomenon may be accounted for thus. When the prong of the 

 tuning-fork beats outwards, the air is driven before it and is con- 

 densed, while behind it there is a partial vacuum. If the velocity 

 -of the prong was greater than that of a particle of air in its free 

 path movement, then there would be a complete vacuum behind 

 the prong. As the latter beats to and fro, it is obvious that the 

 density of the air adjacent to the prong must be less than if the 

 latter was at rest, the difference depending upon the relative ve- 

 locity of the prong to that of the molecules of air in their free path 

 movements. As the pressure of the air varies as its density, it fol- 

 lows that the air-pressure is less in the neighborhood of the vibrat- 

 ing fork than at a distance from it. Hence, if an object is near to 

 the vibrating fork, the air-pressure will be greater on the remote 

 ■side, and will push the object towards the source of vibrations. 



Numerous devices have been invented by Doornak and Strop to 

 .-illustrate this principle. Most of them are too complicated and 

 ■costly to be had by more than a few. The following is simple 

 enough, and can be available for any one having a Chladni plate. 



Cut a disk three or four inches in diameter out of letter-paper, 

 ■and then cut eight or ten radial slits from the circumference half- 

 "way to the centre, and turn up one edge of each sector so as to 

 form a kind of paper windmill. Suspend this by a thread from 



its centre, and see that it hangs horizontally, which may be done 

 by fixing a bit of beeswax to the middle of the disk, and have the 

 thread go through it. Adjustment will be easy and quick by 

 slight pressure upon the wax, changing the relative position of the 

 thread. 



This disk may now be brought over a properly mounted Chladni 

 plate near the edge, and as close to it as possible, while allowing 

 it free space for rotation without touching the plate. If the plate 

 be made to vibrate vigorously, the disk will begin to spin, turning 

 in the same direction as if a current of air were blowing upon it 

 from above. The lower components of the sound of the plate will 

 be necessary to make so large a disk as the above to spin, as the 

 higher ones have too many nodes. The fundamental is the best ; 

 and, if it can be produced with an amplitude of an eighth of an 

 inch or more, the disk will go round two or three times a second. 

 Of course, the bow should be drawn across the edge opposite to 

 the disk, in order to prevent a node being formed underneath it, 

 and also to avoid the disturbance from movements of the air. I 

 have found that the fundamental vibration of the Chladni plate 

 can more easily be produced by bowing it with a round wooden 

 rod well rosined, than with the ordinary violin-bow. In this exper- 

 iment the pressure of the air is lessened between the nodes at the 

 surface of the disk, and the space thus affected extends to the 

 height of an inch or two. It is also evident that the light dust 

 that moves to the place of greatest disturbance is moved there by 

 the difference in air-pressure instead of by little whirlwinds caused 

 by the vibrations, as it was explained by Faraday. 



■ A. E. DOLBEAR. 

 College Hill, Mass., March 5. 



Note on the Robinson Anemometer Constant. 



This is the factor by which the velocity of the central point of 

 the cups is reduced to the actual velocity of wind. When Dr. 

 Robinson first invented his anemometer in i860, he determined the 

 value of this factor, for all patterns of the instrument, to be exactly 

 3, and this has been in use for all patterns ever since. But by the 

 experiments of Dr. Dohrandt at St. Petersburg in 1878, this con- 

 stant, for the Kew pattern at least, was found to be much less, 

 which led Dr. Robinson to repeat his experiments ; and the result 

 was a confirmation of Dr. Dohrandt's result, and showed that his 

 own factor is erroneous. Experiments at the Deutsche Seewarte 

 in Hamburg have also given a factor much smaller. 



The labor of the wind-force committee of the Royal Meteorologi- 

 cal Society, referred to in my previous note {Science, xiii. p. 171), has 

 been directed mostly toward determining this factor for several 

 anemometers of different patterns, which are as follows : — 



These were placed near the end of a long arm of a whirling 

 apparatus, moved by a small steam-engine with varying velocities. 

 The number of turns of the anemometer compared with that of 

 the whirling apparatus during any given time of uniform velocity, 

 the relations between the length of the arm of the anemometer 

 and the distance on the arm from the centre of whirling being 

 known, gave the ratio between the velocities, and so the value of 

 the constant, which is found to be about the same for all, except 

 very small, velocities. 



The average of 58 experiments with the Kew Standard gives 2. 15 

 instead of 3 for the value of this constant. From 51 experiments 

 made with A 19, the value 2.51 was obtained, while the average of 

 49 experiments with A 21 gave 2.96, which is very nearly that de- 

 termined by Dr. Robinson, and now in use. It is seen, therefore, 

 that while the Robinson factor is very erroneous for the Kew pat- 

 tern, and also for A 19, but especially the former, it is very nearly 



