August 31, 1S83.] 



SCIENCE. 



285 



licved to be. This form is convenient because it 

 answers for demonstrating the law at pressures above 

 ami below one atmosphere. Of course, for the latter, 

 the air in X» must be removed through S as at first. 



Natural suoTv-balls or sno'w-roUers. 



BY S.\MrEI, nAKT OF n.\RTFOI!I>, COXX. 



The author carefully described the production and 

 appearance of snow-balls or snow-rollers, the result 

 of natural causes, of which a fine example was pre- 

 sented in Connecticut and southern Jlassachusells 

 last February. Snow had fallen, and had been cov- 

 ered with a frozen surface by a liuht subsequent rain. 

 Upon this surface fresh snow fell, and this under the 

 influence of wind was collected in masses of differing 

 shapes. Some were spherical, from one to nine inches 

 in diameter; most were rollers shaped like muffs, 

 cylindrical, with a conical depression at each end 

 reaching nearly to the middle. The largest observed 

 by the author was IS inches long, and 12 inches in 

 diameter; but some wore reported much larger. The 

 path of formation showed that the roller had started 

 with a small pellet, and, gaining both in length and 

 diameter, had rolled up a long isosceles triangle of 

 snow from its vertex. These paths were observed 

 of a length of 2.5 to 30 feet, and others were reported 

 as of (iO feet. The paths of the round balls were of 

 nearly the same width throughout. None of these 

 masses could be lifted without breakage. Such rollers 

 were seen over an area of 40 miles in length: the 

 author believed that they must have been millions in 

 number. 



Remarks on the tracings of self-registering in- 

 struments, and the value of the signal-service 

 indications for lo-wa, in June and July, 1883. 



BY GUST.WUS mXIUCUS OF IOWA CITY, IO. 



This paper was mainly a severe criticism on the 

 work of the signal-service bureau. The author 

 claimed that the predictions of the weather for Iowa 

 in June had been quite untrustworthy, only 50 per 

 cent proving correct. His views as to the value of 

 this service were vigorously combated in a discussion 

 which followed the reading of his paper. 



A new^ heliostat. 



BY B. F. TnOM.VS OF C0I.UMI5IA, MO. 



The instrument is intended to throw the reflected 

 ray horizontally in the meridian. 



A polar axis A C, driven by clockwork, is provided 

 with a declination-arm E F, with arc, pivoted at C, 

 and pivoted to the longer axis of the mirror at E and 

 Ji'. A 'horizontal axis" ZJ G //, with the shaft por- 

 tion B supported (by proper bearings in the standard 

 not represented) in the same verticil plane with the 

 polar axis, is pivoted to the mirror in its shorter axis 

 at G and H, the points E C F G II being always in 

 one plane, and (! aud // so adjusted that a line 

 through them will pass through C perpendicular to 

 E C F. Moreover, the axis of the shaft portion li, 

 extended, is perpendicular to and bisects G C II at 



C. In order that the reflected ray be thrown hori- 

 zontally in the meridian by our mirror, two condi- 

 tions must be met: viz., 1°. The longer axis of the 

 mirror must make and maintain the proper angle be- 

 tween itself and a line from C to the sun; or, what is 



the same thing, the angle ACE must equal latitude 

 plus one-half the sun's altitude at noon. 2°. The 

 shorter axis must move in a vertical pl.ane, perpen- 

 dicular to the meridian ; or, in other words, the prime 

 vertical. These conditions are rigidly met by the 

 above combination. 



The static telephone. 



BY PBOF. A. E. DOLBEAU OF COLI.KGE IlILI,, M.V8S. 



In the static telephone, a ring of hard rubber is 

 used, within which are two jjarallel metal-plates sepa- 

 rated by a body of air (a non-conductor), one plate 

 connected with the line carrying the current. The 

 electrifying of one plate then causes attraction or 

 repulsion of the free plate, and thus a sound in the 

 receiver. This does away with magnetism. This 

 system therefore requires a very large electro-motive 

 power, and uses an induction coil of 2.000 ohms. A 

 ground or return circuit is not present here. The 

 equivalent is the body itself. There is no passage of 

 electricity from plate to plate: the action is purely 

 inductive through space. The insulation is accom- 

 plished by the intervening air-space, and by a coating 

 of varnish, — an excellent di-electric. There is a de- 

 vice to discharge the induction plate in connection 

 with this instrument, which keeps it constantly up to 

 its full possibilities. When this instrument is fully 

 charged, and the electrical conditions are perfect, the 

 receiver may be entirely disconnected from the trans- 

 milter, and sounds and conversation can still be heard, 

 even across a room. 



lie also called attention to the fact, that instru- 

 ments that have been in use work much better tlian 

 new ones, as each plate acts as a condenser. 



In the discussion which followed. Dr. Dolbear was 

 asked if the state of the atmosphere, in any way, 

 affects the operation of the instrument. I have used 

 those instruments, said he, on an actual line between 

 lioslon aud New York, in a night when it rained over 

 the whole length of the line, and the whole line was 



