540 



NA TURE 



[September 27, 1906 



The Rusting of Iron. 



During the past few months the study of the chemical 

 changes involved in the rusting of iron has been coming 

 to the fore. In 1888 Crum Brown pointed out that iron 

 remained free from rust in an atmosphere of oxygen, 

 carbon dioxide, and water vapour so long as liquid water 

 was prevented from condensing on its surface. Whitney, 

 in 1903, confirmed the opinion that liquid water alone 

 had no effect on the metal at ordinary temperatures. No 

 mention was made, however, of the purity of the iron 

 used. Last year Dunstan, Jowett, and Goulding con- 

 firmed these results for polished iron plate (99-94 per cent, 

 iron) in a series of carefully planned experiments. Since 

 iron of such great purity as this is seldom used for com- 

 mercial purposes, it seemed to me desirable to try the 

 effect of water alone on different samples of varying 

 qualities. Three such were chosen : — (i) cast iron from 

 .1 piece of old piping; (2) wrought iron; (3) fairly pure 

 iron (99-5 per cent.). The pieces were polished, and 

 measured approximately 1 cm. long by 3 cm. broad and 

 02 cm. thick. They were dropped into flasks of boiling 

 distilled water, and after five minutes the latter were 

 <.losed with tightly-fitting india-rubber bungs, in the way 

 indicated by Whitney. It was found that whilst the pure 

 ,ind wrought iron were unchanged, the cast iron invariably 

 lurned a shade darker in tint. The experiment was varied 

 by employing thin glass tubes instead of flasks, and the 

 surface of the metal was in some cases roughened with a 

 coarse file. After the admission of the iron, the tubes 

 were drawn out and finally sealed off. The results were 

 invariably the same. I have kept these tubes for several 

 months, but no further changes have taken place. This 

 seems to indicate that, whilst neither warm nor cold water 

 has any effect upon the purer forms of iron, they exert 

 some slight action on the coarser cast iron. 



Many and various are the theories which have been 

 suggested from time to time to account for the process of 

 rusting. Crum Brown pointed out that carbon dioxide was 

 necessary. This dissolved in the water and attacked the 

 iron, forming ferrous carbonate, FcCO,, or perhaps the 

 soluble ferrous hydrogen carbonate, FeH,(CO,),. The 

 hydrogen gas set free combined with any dissolved oxygen, 

 forming water. The oxygen of the air would convert the 

 ferrous hydrogen carbonate into rust, with the liberation 

 of carbon dioxide. Thus a small amount of carbon dioxide 

 in the presence of water and oxygen would be capable of 

 converting an infinite amount of iron into rust. 



During the present year Moody has confirmed this theory 

 by showing that if elaborate precautions are taken to re- 

 move every trace of carbon dioxide, pure iron (99-98 per 

 cent.) may be kept for an indefinite time in the presence 

 of air and liquid water without undergoing the slightest 

 visible change. He has also directed attention to the fact 

 that when a piece of pure iron is introduced into a dilute 

 solution of distilled hydrogen peroxide the latter is decom- 

 posed slowly, evolving a steady stream of oxygen, whilst 

 the iron is unchanged. This again demonstrates the fact 

 that oxygen and water alone have no action on pure iron. 



I have repeated the experiments with hydrogen peroxide, 

 using the different samples of iron already referred to. 

 The peroxide was from Merck, and guaranteed to be pure. 

 It was diluted to thirty times its volume with freshly-boiled 

 distilled water. On introducing the iron, it was found 

 that the pure sample remained perfectly bright, a slow 

 stream of oxygen being evolved. After some hours an odd 

 speck or two of rust appeared. No further alteration 

 occurred even after the lapse of one or two weeks. The 

 wrought iron decomposed the peroxide rather more rapidly, 

 and the specks of rust were more numerous. The cast 

 iron decomposed the peroxide with astonishing rapidity, 

 and in a few minutes was covered with rust. This was, no 

 doubt, due to catalytic action. 



We thus see that the purer the iron the less is the 

 action of the peroxide upon it. Had such pure iron as that 

 used by Moody been employed, I have no doubt my result 

 would have exactly coincided with his. 



It is not impossible, therefore, that while carbon dioxide, 

 oxygen, and w-ater are essential for the rusting of pure 

 iron, the last two alone may be sufficient to cause rust 

 in the coarser forms, such as cast iron. 



J. Newton Friend. 



The Mixed Transformation of Lagrange's Equations. 



The history of the formula 



L = 3; + 2(0fl) - A- V (I) 



is as follows. About twenty years ago I read two papers 

 by Lord Rayleigh and Prof. W. M. Hicks in which certain 

 problems relating to the motion of a cylinder in a liquid, 

 which possesses cyclic irrotational motion, were solved. 

 Both authors employed the old-fashioned method of calcu- 

 lating the forces due to the pressure of the liquid ; but I 

 at once perceived that some form of Lagrange's equations 

 must exist which would enable the problems to be solved 

 without introducing internal forces. I accordingly examined 

 all the works on dynamics to which I had access, including 

 Dr. Routh's treatises and Prof. J. J. Thomson's recently 

 published papers in the Phil. Trans., 1886 and 1887, but 

 without finding what I wanted. The necessary clue was 

 at length obtained by means of a theorem of Lord Kelvin's 

 published in the Proc. Roy. .Soc. Edin., vol. vii., p. 668 

 (about 1S72 or 1873), which enabled me to establish the 

 formula in question. 



Dr. Routh (" Rigid Dynamics." pp. 319 and 320, fourth 

 edition) has given some rather formidable determinants, by 

 means of which it is conceivable that (i) might be deduced 

 by a more or less lengthy analytical process ; but in their 

 present form I have never been able to make any use of 

 them. The procedure explained in §§ 418-420 could not 

 apparently be employed when the velocities which are to 

 be eliminated are either unknown or would be inconvenient 

 to introduce. A. B. Basset. 



Fledborough Hall, Berks, September 21. 



Suspended Germination of Seeds. 



Some years ago it was reported that charlock seed had 

 germinated upon the site of a Norman church in Kent. 

 Is there any similar record of foxgloves awakening from 

 a long sleep? Last February I removed an ancient wall 

 circling the top of a very bare hill on a north country 

 farm. We took out the large foundation stones. As the 

 spring advanced, the site of the wall became carpeted with 

 seedling foxgloves ; if the cattle permit, a thick foxglove 

 hedge will round the crown of the hill next year. 



There were no foxglove plants within several hundred 

 yards, and even had there been roots there would be no 

 seed in February. The wall was formerly the fence of 

 an oak wood, which was felled and turned into pasture 

 forty years ago. The seeds were unquestionably as old 

 as that date ; but my own strong opinion is that they were 

 right underneath the foundation stones of the wall, and 

 had lain there ever since it was built. 1 examined the site 

 very carefully, and also noted that disturbance of the 

 neighbouring turf, outside the site, did not produce any 

 foxglove crop. I believe that the oaks were planted and 

 fenced by a man named Stephen Green between 1600 and 

 1610. 



Another less pleasing instance occurred on the same 

 farm. I took some cartloads of turf and loam — top-spit 

 from an old pasture traditionally called the " Barley Field " 

 — and spread it on another part of the grass ; whereupon 

 there came up thousands of corn-weeds, such as fumitory 

 and sun spurge, which were previously unknown in the 



pastures. 



H. B. P. 



Optical Illusions on Electric Fan. 



k REVOLVING electric fan with gilt blade is illuminated 

 by the light from a window. When we look fixedly on 

 the revolving face an irregular patch of greyish-purple 

 colour appears on the yellowish ground. The patch shows 

 an amceba-like motion, and its size seems to increase with 

 the speed of the fan. The border of the patch is coloured 

 pale. In its centre a bright spot is often discerned. If 

 we look at the fan after having closed or turned aside our 

 eyes for a while the patch has disappeared, and it takes 

 a few seconds before it reappears. 



Several other experiments on illusion can be made con- 

 veniently by means of the fan. If the blade be covered 

 with red papers and revolved slowly, a white paper looked 

 at through the revolving face appears greenish, and a 

 greenish one greyish. T. Terad.a. 



Physical Laboratory, Tokyo, August 26. 



NO. 1926, VOL. 74] 



