324 



NATURE 



[June 26, 1919 



The author distinguishes three forms of soil 

 water : hygroscopic water, a thin film absorbed 

 from the air and condensed on the particles of 

 the dry soil; capillary water, also a film, but 

 thicker than the preceding, taken up by soil in 

 contact with liquid water and held by surface 

 forces ; and gravitational or free water, which can 

 drain away, and, indeed, should be allowed to do 

 so wherever it assumes unduly large proportions. 

 No mention seems to be made of the mole plough, 

 which, in this country, has proved of great value 

 in drainage work. Afterwards comes a chapter 

 on the bacteria of the soil, followed by one on 

 soil air and soil temperature. 



The remainder of the book deals with fertilisers. 

 It is evident that American farmers suffered 

 much less from shortage of fertilisers than did our 

 own farmers as the result of the war. For, whilst 

 an English book written, like the book before us, 

 in 1918 would have been compelled to devote much 

 space to substitutes and to revise considerably the 

 descriptions of processes and comparative stand- 

 ards, the author did not find such alterations 

 necessary, and his chapters differ little from 

 what might have been written before the war, 



LETTERS TO THE EDITOR. 



[The Sditor does not hold himself responsible for opinions ex- 

 pressed by his correspondents. Neither can he undertake to 

 return, or to correspond with the writers of, rejected manu- 

 scripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications.] 



The Credibility of Long-continued Experiments. 



At the Rothamsted Experimental Station certain 

 experiments are continued for a long series of years 

 in order to amass sufificient data to allow of proper 

 statistical treatment. Some of the experiments have 

 been carried on since 1843, others since 1852, 1856, 

 i860, etc. A characteristic feature of the work is 

 the length of time for which particular observers are 

 responsible for their records, some being in charge 

 of the same work for twenty, thirty, or forty years. 

 The fundamental weakness in such long-continued ex- 

 periments is one inherent in human nature itself : 

 errors once introduced are apt to persist, and to cause 

 much harm unless they are soon detected. 



In many of the experiments it is not possible to 

 institute any very satisfactory check on the results. 

 In some cases, however, this can be done. The 

 measurements taken at the rain and drain gauges 

 afford an instance, and the agreement is so close as to 

 deserve record. 



In 1870 three drainage gauges and one rain gauge 

 were set up at Rothamsted, each i/iooo acre in area. 

 The rain gauge is simply a very large funnel embedded 

 in the soil ; the drain gauges are also large funnels, 

 but filled with soil to depths of 20, 40, and 60 in. 

 respectively. In constructing them, however, the soil 

 was not disturbed, but was left in its natural position, 

 whilst the framework of the gauge, by an ingenious 

 arrangement, was built round it. From 1870 to the 

 present day readings have been taken of the amounts 

 of water percolating through the drain gauges : this 

 amount being some 40 to 60 {>er cent, of the water 

 collecting in the rain-gauge. Samples of the water 

 from each gauge are then sent to the laboratory, 

 where the chlorine and nitric and ammoniacal nitro- 

 gen are^ determined. 



NO. 2591, VOL. 103] 



This work went on without intermission from 188S 

 until 1916, when it was suspended owing to the. 

 sudden death of Dr. N. H. J. Miller, who had been 

 in charge the whole of the time. 



The results have now been calculated out and 

 added up. 



It is well established that soil neither absorbs nor 

 gives up chlorine to water containing sodium chloride 

 in solution, therefore the amounts of chlorine found 

 in the drain gauges ought to be equal to that in the 

 rain gauge if the numerous separate records were 

 accurate. Over a short period there is always liable 

 to be a difference, because some of the chlorine may 

 not yet have had time to percolate, but over a long 

 period this is eliminated. The actual results obtained 

 at Rothamsted are :— 



Chlorine in lb. per acre. 



From drain gauges 



Average per 

 annum for 

 4 years 

 1882-92 

 1892-96 

 I 896- I 900 

 1900-04 

 1904-08 

 1908-12 

 I9I2-16 



. i»»»-i9ib 



The number of measurements involved is very large ; 

 there are some 18,000 readings at the gauges and a 

 large number of titrations in the laboratoi-y. The 

 gauge-reading has to be multiplied by its titration 

 value, and the resulting figures are then added up. 

 Considering the multiplicity of the data, the agree- 

 ment in the results is remarkable ; the widest 

 divergence over twenty-eight years is only 2 per cent. 



This close agreement is the result of careful daily 

 work, and not of accident. There is no correspond- 

 ence in the laboratory between the rain and the 

 drainage samples; this is prevented by the cairy-over 

 of water and of chlorine in the drain-gauges from 

 day to day, and even month to month. Nor is there 

 any possibility of straining readings to compel agree- 

 ment; the figures were not regularly added up during 

 the course of the work, but only at rare intervals. 



The result shows how accurately continuous ob- 

 servations can be made provided care is taken. The 

 readings at the gauges have throughout been taken 

 by Mr. E. Grey, who without fail and in all weathers 

 has stuck to the work. The titrations were made by 

 the late Dr. Miller, who would have felt great pride 

 in the final result had he lived to see it. Fortunately, 

 Mr. Grey is still in charge of the gauges. 



E. J. Russell. 



Rothamsted Experimental Station, Harpenden. 



The Lustre of Some Feathers of Humming Birds. 



The brilliantly metallic feathers of the crests and 

 gorgets of most humming birds, which are also 

 erectile, must have a great significance, and present 

 an interesting problem. Why should the most intense 

 brilliancv be on those particular spots? 



An explanation suggested itself while observing a 

 doctor examining the throat of his patient, in bright 

 sunshine, bv the help of a laryngoscope fixed upon 

 his forehead, his patient being placed with his back t 

 to the light. ^ _ i|| 



Holding a humming bird, in bright sunshine, irr IJ 

 front of the corollas of flowers that were turned away || 

 from the light, the illumination of the inside of the^ 

 corolla was most striking, and its use in revealing any 

 small Insects it might contain became quite apparent^ 



