April 6, 191 1] 



NATURE 



179 



iron could only have condensed on the inside of ihe tube 

 in which the metal was contained. 



Oxygen, made by electrolysis of pure baryta solution, 

 was hnally allowed to enter the vessels. The paraffin wax 

 used to hne the quartz and glass vessels was purified by 

 boiling with many changes of " conductivity " water, with 

 frequent shaking, for about loo hours. The iron used 

 was a commercial electrolytic sheet iron containing 99-9 

 per cent. iron. It was carefully cleaned with finely divided 

 carborundum. 



The results were the same in the quartz vessels and in 

 the quartz and glass vessels lined with paraffin wax. 

 Rusting was visible in all cases within a few hours. 



It was only when pure iron, made as described in the 

 paper, was used that no rusting took place. 



.\fter these and other experiments, it was considered 

 that quartz vessels, which lend themselves admirably to 

 cleaning and purification, were the most suitable vessels 

 in which to study the reaction. 



It is claimed by the authors that they have established 

 the following facts : — (i) Pure iron does not rust when 

 placed in contact with pure water and pure oxygen. (The 

 term pure is used, of course, in a relative sense.) (2) The 

 contention that " the dominant factor in the atmospheric 

 corrosion of commercial iron is carbonic acid," or any 

 other acid, is untrue. This must not be taken to mean, 

 however, that carbonic acid plays no part in the atmo- 

 spheric corrosion of ordinary commercial iron. 



Bertram Lambert. 



Chemical Department, University Museum, Oxford. 



The reference made to Mr. Lambert's work in my note 

 on the rusting of iron has already proved of value in 

 eliciting from the author a statement in reference to tests 

 carried out in vessels lined with purified paraffin wax. 

 His statement does not, however, provide a solution of 

 the problem to which attention was directed, namely, that 

 commercial iron exposed freely to air and water under the 

 conditions described by Moody and by Friend does not 

 rust, whereas in his own experiments rusting took place 

 in the case of all but the most highly purified samples. 

 It is difficult to predict whether the essential difference 

 Nvhich still awaits discovery will ultimately be found in 

 I he iron, in the air, in the water, or in the vessel, but it 

 - to be hoped that further discussion — in the columns of 

 Nature or elsewhere — may speedily throw light on this 



tlicult and important matter. 



It is a common experience, of which Mr. Lambert's 

 \\f>rk has provided excellent illustrations, that an un- 

 suspected impurity (such as platinum in iron) is far more 

 likely to promote than to prevent chemical change. It 

 was for this reason that attention was directed first to the 

 experiments in which rusting actually' took place rather 

 than to those in which it was successfully prevented. My 

 suggestion in reference to silicic acid is correctly described, 

 but I do not think that a note of exclamation was called 

 for in view of the fact that mere contact with glass, as 

 Moody found, actually produces the effects referred to, not 

 in oxygen, but in purified air. The quantity of acid re- 

 quired to initiate the rusting must be extremely small, 

 but its complete removal seems to outweigh in importance 

 every other factor that may be taken into consideration. 

 This view is amply supported by the fact that Friend, 

 using common iron, common air, and the simplest 

 apparatus, was able to stop the rusting by the simple 

 device of washing the iron with a common alkali and then 

 rinsing it with common water evaporated from the alkali 

 and condensed on the iron. 



With the first of Mr. Lambert's claims I am entirely 

 in agreement ; his second claim I regard as unproved and 

 fallacious. To the statement that " the dominant factor 

 in the atmospheric corrosion of commercial iron is carbonic 

 acid " I adhere unreservedly, not only because the corrosion 

 can be stopped by the addition of an alkali or by the 

 removal of acid, but even more emphatically because the 

 carbonic acid leaves behind an irrefutable proof of its 

 activity in the presence of something like 10 per c^nt. of 

 ferrous carbonate in the rust formed under the common- 

 place conditions to which I referred. T. M. L. 



NO. 2162, VOL. 86] 



Apiil MeieoTS. 



The following are the most important meteoric events 

 that become due from about April 6 to April 30 in the 

 present year : — • 



Epoch April 5, 2ih. (G.M.T.), approximately second 

 order of magnitude. Principal maximum April 6, 

 2ih. 5Sm. ; secondary maximum April 6, ih. 



Epoch April 7, lyh, eighteenth order of magnitude. 

 Principal maximum April 9, gh. 40m. ; secondary maxima 

 April 8, 2h. 15m., April 8, iih. 30m., and April 9, 

 2ih. 30m. 



Epoch .^pril 10, lyh., twenty-third order of magnitude. 

 Principal maximum April 9, oh. 15m. ; secondary maxima 

 .April 8, 2ih. 15m. and April 10, i4h. 30m. 



Epoch April 11, 4h., twenty-eighth order of magnitude. 

 Principal maximum April 11, iih. 15m.; secondary maxi- 

 mum April 12, yh. 15m. 



Epoch April 13, 22h. 30m., thirty-third order of magni- 

 tude. Principal maximum April 14, I5h. 40m.; secondary 

 maximum April 15, i4h. 40m. 



Epoch April 17, lyh., twenty-ninth order of magnitude. 

 Principal maximum April 18, 2h. 15m. ; secondary maxi- 

 mum April 18, 4h. 50m. 



Ep>och April 19, 8h., twenty-fourth order of magnitude. 

 Principal maximum .April 17, i3h. 40m. ; secondary maxi- 

 mum April 17, 2ih. 45m. 



Lyrid epoch April 19, i4h. 30m., approximately eighth 

 order of magnitude. Principal maxima April 17, 

 2oh. 50m., and April 19, qh. 50m.; secondary maxima 

 .April ic), 9h. 30m., and April 19, iih. 40m. 



Epoch April 23, 3h. lom., approximately sixth order of 

 magnitude. Principal maximum April 21, i6h. 40m.; 

 secondary maxima April 21, 3h. S5m., and .April 22, 

 2oh. 30m. 



Epoch April 25, 6h. 30m., eleventh order of magnitude. 

 Principal maximum April 23, i3h. ; secondary maxima 

 April 23, 23h. 30m., and April 24, 22h. 50m. 



Epoch April 26, 6h., thirty-first order of magnitude. 

 Principal maximum April 27, i2h. lom. ; secondary maxi- 

 mum April 27, 4h. 40m. 



Epoch April 27, ih., approximately tenth order of 

 magnitudf^. Principal maximum April 28, 22h. 50m. ; 

 secondary maximum .April 27, 23h. 50m. 



Shooting stars should be numerous on April 19, as 

 several maxima occur shortly before midnight on this 

 date, among which Lyrids and bright meteors from allied 

 radiants will probably be strongly in evidence. There is 

 also another Lyrid display on the night of April 21, but 

 it will only be partially visible on this side of the Atlantic. 



April 3. John R. Henry. 



Insurance against Rain. 



.As this scheme of insurance is one of the few of which 

 data can be obtained by " the man in the street '* from 

 which the odds in favour of the underwriters can be 

 calculated, I went to the trouble of investigating the 

 matter. Possibly the results may interest your readers. 



In " British Rainfall " for 1909, by Dr. Hugh Robert 

 Mill, statistics are given of the numlx^r of days on which 

 varying quantities of rain fell. The figures give the aver- 

 age for thirty-seven rainfall stations in England and Wales 

 over a period of seven years, 1903-9. From these, by a 

 method of interpolation, it is found that fifty-eight days 

 in a vear may be expected to show a fall of 020 of an 

 inch or more, and seventy-four of 015 or more. 



Working out the probabilities on the basis of these 

 figures, assuming that rain on one day is independent of 

 rain on any other day, the following results are 

 obtained : — for every looZ. received in premiums the under- 

 writers may expect to have to pay, in the case of Policy A, 

 66/. ; Policy B, 74/. ; Policy C, 43/. ; and Policy D, 64/. 



In favour of the underwriters is the fact that in the 

 south and south-east, where alone the scheme is applic- 

 able, the probability of rain is less than in England and 

 Wales generally. A further point in their favour is that 

 the summer months arc less wet than the mean of the 

 whole year on which the above figures are calculated. 



C. O. Bartrum. 



32 Willoughby Road, Hampstead, N.W., March 30. 



