NA TURE 



[May 14, 1Q08 



do not support the conclusion that the activity of these 

 salts is a normal atomic property of potassium, and that it 

 is always directly proportional to the amount of that metal 

 present in the salt. 



In measuring and comparing the activities of the 

 different salts, these were spread in turn in thin regular 

 layers on a shallow tray, which was placed on the bottom 

 of an ionising chamber 40 cm. long, 26 cm. wide, and 

 28 cm. deep. The saturation currents through the air in 

 this chamber were measured with a sensitive quadrant 

 electrometer, and were taken as measures of the activities 

 of the different salts. Experiment showed that the satura- 

 tion current increased with the thickness of the salt layer 

 up to between 2 mm. and 3 mm., and for greater thick- 

 nesses remained constant. 



The investigation included the examination of some 

 thirty specimens of potassium salts, and thirteen samples 

 of the salts of the other alkali metals. As a result of this 

 examination it was found : — 



(i) That samples of a selected potassium salt obtained 

 from different sources exhibited widely differing degrees 

 of activity. Two chlorides of potassium, for example, 

 were found to differ by more than 40 per cent, in their 

 activities, and two of hydroxides by an almost equal 

 amount. In the case of cyanide of potassium, the varia- 

 tion in activity was especially marked, as two samples of 

 this salt exhibited activities which were approximately 

 only 5 per cent, and 20 per cent, respectively of that shown 

 by a number of other samples of the same composition. 

 A sample of potassium sulphite, too, was found to possess 

 an extremely small activity. 



With the majority of the salts, however, the variations 

 were not so marked, but the differences observed, even 

 when due allowance was made for the varying densities 

 and states of division of the salts, were so extensive and 

 of such magnitude as practically to preclude the view th.nt 

 the activity of potassium and its salts was connected with 

 a normal atomic property of the metal. 



(2) That while metallic sodium and several sodium s.ilts 

 did not exhibit the slightest trace of activity, some samples 

 of sodium chloride, obtained in the form of rock salt, 

 showed an activity comparable with that exhibited by a 

 number of the potassium salts. This result, taken in con- 

 junction with the low value obtained by Elster and Geitel 

 in their measurements on the conductivitv of air in a salt 

 mine, would indicate that very probably some active 

 impurity was present in the samples of rock salt examined. 



(3) That with the exception of ammonium chloride, 

 which exhibited a feeble activity, none of the lithium and 

 ammonium salts examined showed the slightest trace of 

 radio-activity ; that a sample of rubidium alum was found 

 on examination to exhibit an extremely small activity, and 

 that a sample of ca-sium chloride exhibited one which was 

 only just measurable. J. C. McLennan. 



University of Toronto, .April 15. 



Chemical Analyses of Water from Dew Ponds. 



I H.iVE been interested in the reviews of books and 

 articles which have appeared in Nature from time to time 

 on the subject of dew ponds ; and it occurred to me that 

 the chemical analysis of the water of these ponds would 

 help to settle the question of the origin of the water. This 

 is a method used to some extent by sanitary authorities. 

 I have had the opportunity of obtaining some specimens 

 of water from different districts, and the specimens have 

 been analysed by Mr. Claude Saville Grace, one of the 

 students at this institute. 



The first specimen (o) came from a dew pond on the 

 southward down to the north of Ramsbury, Wiltshire. 

 The pond is on the flat upland near the 6c)3-feet mark on 

 the i-inch Ordnance map south of Aldbourne. The forma- 

 tion is chalk, so that analysis would immediately settle the 

 point as to whether the water had come through the chalk 

 or had been condensed from the air. The other two speci- 

 mens, (b), (c), come from St. Boniface Down, north of 

 Ventnor, Isle of Wight. The pond (b) is near the 7S7-feet 

 mark, almost at the highest point of the down ; the second 

 pond (c) is on the neck between St. Boniface and Shanklin 

 Downs. The downs are chalk masses lying on.Greensand 

 rock. I have added the analysis of the St. Boniface spring 



NO, 20 II, VOL, 78] 



water, a spring on the south side of the down facing 

 Ventnor, about 450 feet above sea-level. It is locally known 

 as a wishing well, and its chemical peculiarity is that it 

 contains sulphuretted hydrogen in small quantity. It un- 

 doubtedly comes out of the chalk, and the sulphuretted 

 hydrogen is due to the decomposition of pyrites which 

 occurs in masses in the chalk. 



(ff) Ramsbury pond 



{/>) St. Boniface pond .., 



(1) Shanklin pond , 



((/) St. Boniface Wishing Well. 23-9 



The quantities of CaCO^ indicate to me that the waters 

 are, in the cases (n), (b), (c), condensed waters which have 

 been lying in the pond sufficiently long to take up a little 

 CaCO,. The ponds near the sea show increased chlorine, 

 probably from salt spray blown up from the sia. The 

 sulphuretted hydrogen shows the origin of the wishing- 

 well water in the deep chalk. 



To me there is very little doubt that all three ponds are 

 simply water butts in which rain water is stored. In- 

 spection of the ponds shows that they have much larger 

 catchment areas than simply the water area, and the area 

 is generally more than nine times that of the pond. We 

 have to remember that for circular ponds and catchment 

 areas of radii i and 3 respectively a rainfall of 20 inches 

 would mean a depth of 180 inches (15 feet) when collected 

 into the pond area, so we can easily understand the 

 presence of water all the year round under these con- 

 ditions. 



It is interesting to note that a fresh-water pond is easily 

 distinguished from the hard-water pond by the waterweed 

 growing in the former. I have noticed the same carex in 

 all the fresh-water ponds. Sidney Skinner. 



South-Westcrn Polytechnic Institute, May S. 



The Reflection of Distant Lights on the Clouds. 



I DO not know whether observations have ever b'-en made 

 to determine how far the reflection of distant lights on 

 the clouds may be seen. It may possibly, howev.jr, be of 

 some interest to know that the lights of London may at 

 times be seen in this way at a distance of at least fifty 

 miles. At 11 p.m. on April 30 the reflections of the lights 

 of several neighbouring towns were unusually bright as 

 seen from here. The altitude of the Portsmouth glare was 

 about 10° ; the distance of the centre of Portsmouth is 

 about 12-5 miles ; the cloud height was therefore about 

 2-2 miles. Over Hindhead and Blackdown a bright band 

 of light was visible. Circumstances prevented me from 

 measuring its altitude, but I estimated it as one or two 

 degrees. Now London lies exactly in this direction, and 

 fifty miles would bring one to the well-lighted area of 

 south London. If the cloud height were uniform, the 

 altitude of the reflection at this distance should have been 

 a little more than 2°. The only other large town in the 

 same direction is Guildford ; the altitude of its glare should 

 have been 5°. I do not think I could have made so large 

 an error in estimating the altitude, but apart from this 

 the Guildford glare would not stretch along the horizon 

 for more than 2°, while the observed band of light sti-etchcd 

 for at least 10°, and possibly more, for trees bounded the 

 view to the west and the downs to the east. 



Ch.\rles J. P. Cave. 



Ditcham Park, Petersfield. May 9. 



Jupiter's Eighth Satellite. 



The discovery at Greenwich Observatory of Jupiter's 

 eighth satellite, its great distance from the planet, and its 

 retrogr.ade motion, have excited the interest of the astro- 

 nomical world. 



Until more extended observations have led to a more 

 certain knowledge of the orbit, speculation is premature. 

 But it is impossible to resist the conjecture that there is a 

 bare possibility that the object is really the long lost 

 Lexell's comet, which in 1770 was describing an elliptic 

 orbit with an eccentricity of 0-7858, with a periodic time 



