August 21, 1908] 



SCIENCE 



255 



but best when one electrode is either copper or 

 silver, the arc burning in illuminating gas. The 

 oscillations are most intense with nigh voltage 

 and a current between one ampere and the point 

 at which the arc changes to a glow. To gain an 

 idea of the maximum potential difference across 

 the arc during one cycle an apparatus for obtain- 

 ing a point discharge in hydrogen was connected 

 in parallel with the arc. The results indicate 

 that the effect is an exceedingly rapid change 

 back and forth between arc and glow discharge, 

 of a frequency of something like a million per 

 second. Experiments were described on the use 

 of a resonating circuit near the arc, and on the 

 connection of a capacity and self-inductance 

 across the are. 

 A Study of Overcast Skies: Edward L. Nichols, 



Cornell University. 



The spectrophotometric measurements which 

 formed the basis of this paper were carried out 

 by means of a portable apparatus which could 

 readily be set up during the travels of the author 

 in Europe, and gave the opportunity to compare 

 the skies of widely different localities and dif- 

 ferent times of day. The relative intensities of 

 the different color-components were very different 

 with different kinds of sky. The radiation was 

 scarcely ever selective but almost always of the 

 " black body " type. There is, however, almost 

 always an absorption band in the violet during 

 the middle of a bright day in mountainous re- 

 gions. Its development is coincident with the 

 gathering of a slight mist. 



The illumination from an unclouded sky is 

 about the same as from a completely clouded sky. 

 More light comes, however, from a sky which is 

 partly covered with clouds than from either. 

 The so-called " cumulus " clouds produce espe- 

 cially good luminosity. 

 Demonstration of WiUon's Cloud Experiment, 



etc.: 6. F. Huxl, Dartmouth College. 



This consisted of a demonstration of several im- 

 proved forms of lecture experiments, as follows: 



1. A very simple and satisfactory form of Wil- 

 son's " cloud " experiment, showing the conditions 

 under which the cloud is obtained. 



2. Wehnelt's tube, in which electrons sent out 

 by a " button " of calcium compound in an ex- 

 hausted tube were shown experimentally to be 

 present. 



3. An improvement on Mayer's floating magnet 

 experiment, consisting of passing a single turn of 

 wire around the circular glass vessel, in which 

 the small vertical magnets are made to float. By 



sending current through this wire the effect of 

 surface tension at the boundary is entirely elim- 

 inated. The stable configurations of magnets are 

 different from those secured when a central con- 

 trol magnet is used. But the fields due to the 

 current and to a central magnet are easily super- 

 posed. 



Interest in the floating magnet experiment has 

 been recently revived by J. J. Thomson's theories 

 regarding the structure of atoms. 



The Influence of Temperature on the Fluorescence 

 of Uranium Glass: R. C. Gibbs, Cornell Uni- 

 versity. 



The glass used was a block of canary glass. 

 The light from a mercury arc was sent through 

 this glass, whose temperature was changed in a 

 measurable way through about 400 degrees 

 Centigrade. Results will be published in the 

 Physical Review. 



Some Electrical Properties of Silicon: Feances 



G. Wick, Cornell University. 



The silicon used was made at Niagara Falls. 

 The following properties were noted: 



1. In its resistance it was found to change with 

 temperature as carbon does and not as the metals. 



2. The thermo-electric power with respect to 

 lead was found to be 220, which is larger than 

 for any other substances except selenium and 

 tellurium. 



3. It was found also that, as with selenium and 

 tellurium, the Hall effect was large in silicon. 



4. Rods of the element were used as one ele- 

 ment in voltaic cells, but without developing any 

 special promise of usefulness as compared with 

 copper, etc. In its position in the electromotive 

 series it is very near copper. 



A Study of Short-time Phosphorescence: C. W. 



Waqgenee, Cornell University. 



All good measurements since the time of 

 Becquerel show about the same kind of lumin- 

 osity-decay curve. The recent work which forms 

 the basis of this paper was designed to see if 

 this form is characteristic of substances which 

 lose their luminosity in an exceedingly short in- 

 terval, say within one seven-hundredth of a sec- 

 ond. The apparatus used was that devised by 

 Nichols and Merritt a year or two ago. It was 

 found that practically the same form of decay 

 curve characterizes substances of this class as 

 those whose period of decay is longer. 



An Experimental Study of the Recovery of 

 Selenium Cells: L. S. McDowell, Cornell Uni- 

 versity. (Read by title.) 



