430 



NATURE 



[March i, 1888 



To test this point the following experiment was therefore made. 

 A little dusty air was mixed with filtered air, and put into the 

 test receiver, and saturated with water vapour. An expansion 

 of only 2 c.c. was made; this caused the formation of a fog. 

 After these fog particles had settled, the air was returned to the 

 receiver ; and after a short time another 2 c.c. expansion was 

 made, when other fog particles appeared. After this had been 

 done a number of times, the density of the fog got less and less, 

 and at last entirely ceased. After this an expansion of 5 c.c, was 

 made ; this produced a rainy condensation in the receiver, which 

 appeai'ed a number of times on successive expansions being made, 

 getting less and less dense, and at last it also ceased entu-ely. 

 After all condensation had stopped with the 5 c.c. expansion, the 

 expansion was increased to 10 c.c, when another shower made its 

 appearance, and after one or two expansions all condensation 

 again ceased. After this condition was attained, an expansion of 

 150 c.c. was made with the pump, when scarcely one drop made 

 its appearance. 



It is concluded that in the above experiment we have distinct 

 evidence that the size of the particle does aftect the degree of 

 super?aturation required to produce condensation on it. Because, 

 though an expansion of 2 c.c. produced a supersaturation sufficient 

 •o cause more than one-half of the particles to become visible, 

 yet it required a higher degree of supersaturation to cause con- 

 densation to take place on others. It is also concluded from 

 the experiment that the failure of the air to keep clear, in the 

 tests where high supersaturation was used, was not due to the 

 presence of extremely small particles, as an expansion of 10 c.c. 

 is practically great enough to produce a supersaturation sufficient 

 to cause condensation on the smallest particles. 



The failures in the tests not being due to the presence of 

 extremely small particles, it is concluded that they are true cases 

 of condensation without nuclei, similar to those referred to in a 

 previous communication. It was thought that, if they were true 

 cases cf spontaneous condensation, they might be checked 

 if the expansion was made slowly and free from shocks. And 

 on the other hand any shock would tend to produce condensation 

 in dust-free air if highly supersaturated. On trying this, it was 

 found that no condensation took place if the stroke of the pump 

 was made slowly and steadily, and that if done quickly, and the 

 piston made to strike the cover of the cylinder violently, then 

 copious showers were always produced in the dust-free air. Here, 

 then, was the key to one of the difficulties, and accounted for the 

 occasional increase in the number of the particles counted ; many 

 of the drops having no dust-nucleus. Failure from this cause is 

 now entirely prevented by causing the air on its passage from the 

 receiver to the pump to pass through a small opening, or better 

 through a small cotton-wool filter ; this checks all violent rush of 

 air, and shocks, and keeps the filtered air perfectly free from 

 condensation even when highly supersaturated. 



Again, the failure of perfectly filtered air to keep free from 

 condensation was frequently observed after the inside of the 

 test receiver had been newly wetted. It looked as if the newly 

 wefted sides had saturated the air more thoroughly, and that 

 the condensation was due to the higher degree of supersaturation 

 which took place when expansion was made. This cla^s of 

 failures was, however, traced to the manner of wetting the inside 

 of the receiver. If it was done roughly, and the water splashed, 

 then many nuclei were manufactured in the receiver ; if it was 

 done quietly, none, and no condensation followed. Another 

 cause of failure was traced to a drop of water getting into the 

 pipe by which the air entered, and the inrush of air tearing the 

 water into fine spray, which became active centres of conden- 

 sation. 



As yet no great number of tests of air have been made under 

 different conditions, natural or artificial ; but in the following 

 table will be found sonie of the results obtained by this method 

 of counting. 



Dust Particles in the AttHosphere. 



Source of the Air. _ Numbsr per c.c. Number per c. in. 



Outside Air — Raining ... 32,000 ... 521,000 



), „ — Fair ... i ;o,ooo ... 2,119,000 



Room 1,860,000 ... 30,318,000 



,, near ceiling ... 5,420,000 ... 88,346,000 



Btmsea Flame ... 30,000,000 ... 489,000,000 



In the first column of the table 'is entered the source of the 

 air ; in the second, the number of particles per cubic centimetre ; 

 and, for the benefit of those who think in English measures, the 

 number per cubic inch is entered in the third column. The first 



number in the table, for ordiny.ry outside air, was obtained on 

 January 25, after a wet night. The number given for fair weather 

 is an average got when the weather was in that condition. As 

 yet far too few measurements have been made to enable us to 

 trace any connection between the number of particles and the 

 weather, but it is hoped that something practical may result from 

 observations of this kind. The first number given for the air in 

 a room is the number counted in the air of a room where gas 

 was burning, and taken at a height of 4 feet from the floor ; the 

 other number was counted in air drawn from near the ceiling ; and 

 the last number was got in the air collected over a bunsen flame. 

 The value of numbers given in the table has not been carefully 

 considered, and they are not given as absolutely correct ; great 

 accuracy, indeed, does not seem possible, when we consider the 

 conditions ; and, further, the number is constantly varying. 

 For this reason it has not been thought necessary to make any 

 corrections for temperature and pressure. Though we can get 

 with a fair degree of accuracy the number of particles in the air 

 in the test receiver, yet in all probability the calculated numbers 

 given in the table are rather under than over estimates, as it is 

 difficult to manipulate air without losing much of its dust. For 

 instance, in one hour about one-half of the particles settle out of 

 the air in the gasometer. Though the numbers do seem very 

 large, yet so far as can be judged at present they are fairly correct, 

 and at least represent the kind of numbers we have to deal with. 

 It does seem strange that there may be as many dust particles in 

 I cubic inch of the air of a room at night when the gas is burning'; 

 as there are inhabitants in Great Britain, and that in 3 cubic inches 

 of the gases from a bunsen flame there are as many particles as 

 there are inhabitants in the world. 



John Aitken. 



UNIVERSITY AND EDUCATIONAL 

 IN TELL IGENCE. 



Cambridge. — The Frank Smart Studentship of Botany, 

 founded at Gonville and Caius College by Mr. F. G. Smart, 

 M.A., M. B., and Mrs. Smart, by the transfer of £2i,OQ Great 

 Eastern Railway 4 per cent. Debentures, is to be awarded for 

 the first time at the beginning of Easter Term. Candidates are 

 to send in their names to the Master of the College, Dr. Ferrers, 

 on or before March 20. The electors are the governing body of 

 the College, acting after consultation with the Professor and the 

 Reader of Botany for the time being in the University. The 

 Studentship is to be open to all members of the University who 

 have taken honours in the first part of the Natural Sciences 

 Tripos, and of not more than five years' standing ; but the 

 elected Student must become a member of Gonville and Caius. 

 No competitive, examination is to be held for awarding the 

 iStudentship. The Student is to apply himself to original in- 

 vestigation in botany, and must be able to show that he is doing 

 so at any lime, on penalty of forfeiting the Studentship. The 

 Studentship is to last two years, but may under special circum- 

 stances be prolonged for one year more. The regulations of the 

 Studentship ai"e only to be clianged, after the death of Mr. an(: 

 Mrs. Smart, by consent of the Council of the Linnean Society 

 of London. A prize of £,(i in books is to be given out of the 

 interest of the fund to the undergraduate student of Caius 

 College who shall distinguish himself most in botany at fhe 

 annual College examination. 



The collection of British birds' eggs made by the late Mr. J. 

 P. Wilmot, of Trinity College, containing a specimen of the 

 great auk's egg, and other specimen s figured in Hewitson's 

 " British Oology," has been presented to the University by 

 Lady Caroline and Mr. C. H. Russell, in memory of Mr. George 

 Lake Russell, Lady Caroline's late husband. 



Plans are submitted for the proposed new plant-houses at the 

 Botanic Garden, to cost ^2760, and of a laboratory in the 

 garden, to cost ^'i.i^o. 



SOCIETIES AND ACADEMIES. 

 London. 



Royal Society, February 9. — " The Small Free Vibration- 

 and Deformation of a Thin Elastic Shell." By A. E, H. Love, 

 B. A. , Fellow of St. John's College, Cambridge. 



In this paper the method employed by Kirchhoff and Clebsch 

 lor the treatment of a thin plane plate is applied to the case of a 



