May io, 1883] 



NATURE 



29 



thirds of its time, and utilising also, as I have above explained, 

 the more rapid of the descending currents, the bird can more 

 than sustain itself. It can at will glide to windward at the rate 

 of fifteen miles an hour against the breeze, losing of elevation 

 only one in twenty. R. Courtenay 



L'Ermitage, Hyeres (Var), France, April 28 



Flight of Crows 



I CAN corroborate the observation of Mr. Murphy as to the 

 oblique flight of crows. When I have seen them so flying theie 

 has always been a cross current, and they have merely kept their 

 heads a little to the wind. 



Cambridge Thos. McKenny Hughes 



Sheet Lightning 



Du choc des opinions jaillit la virile. I still adhere to your 

 assertion that sheet lightning is not, at least in most cases, the 

 mere reflection of a common but distant storm. On the high- 

 lands of Ethiopia, in the years 1S42 to 184S I was diligently 

 engaged in investigating the electrical phenomena so frequent in 

 that region. The details of my observations were printed in 

 1858 by the French Institute, and I have published again my 

 results in my "Observations relatives a la Physique du Globe" 

 (Paris, 1873). The following cases may be of interest : — 



Near the zenith eight successive flashes of lightning were seen 

 21 seconds before their thunder, which lasted exactly 12 seconds. 

 Another day it lasted 24 "4s. thirty successive times, and, as pre- 

 viously, without any rain. My greatest observed interval was 

 m - 2s., corresponding to a distance of 38,500 metres, &c. 



I have seen more than once straight or zigzag lightning unac- 

 companied by thunder. One afternoon it went to and fro twice 

 between two horizontal cloud banks, and ended in sheet light- 

 ning which illuminated, not the loner dark bank, but only the 

 under surface of the upper cloud. 1 have observed frequently 

 thunder without lightning and lightning without thunder. 



When in Adwa I recorded silent sheet lightning towards 

 Gondar, 240 kilometres distant, where a violent storm was 

 raging at the same time. Before leaping to a hasty conclusion, 

 let us hear a case bearing pointedly to the opposite opinion : in 

 1845, at Saga (latitude 8° u'), a semi-transparent fog which had 

 mantled over the valley, and could not be more than 3500 metres 

 distant, gave out a flash of sheet lightning without thunder. 



Although my numerous observations have given me a strong 

 bias in favour of your opinion, I do not wish to impose it on 

 reluctant philosophers, but suggest the following system to clear 

 up the question : — Let two observers, A and B, 40 or 50 miles 

 asunder, mention instances of lightning seen in each other's true 

 bearing. If they can also secure the help of a third observer 

 located on or near the straight line from A to B, and who can 

 watch in two opposite directions, many important results may be 

 obtained. Antoine d'Abbadie 



Paris, May 5 



The American Trotting-Horse 



Mr. Brewer's memoir on the evolution of the breed of the 

 American trotting-horse (Nature, vol. xxvii. p, 609), and the 

 statistical tables that accompany it, are full ol interest, but 

 I only propose now to concern myself with the latter, which 

 may be easily and usefully discussed by employing a statistical 

 method that I have long advocated. In explanation I will begin 

 by extracting the final terms of four of the lines of his table, as 

 follows : — 



The meaning of these entries are, that in the year 1871 there 

 were 99 horses that could trot a mile in 2 minutes 27 seconds, or 

 less ; that in the same year there were 40 that could trot it in 

 2 minutes 25 seconds, or less ; and so on. Their significance is 



that the rate per mile of the hundred fastest American trotting- 

 horses has become 2 seconds faster in each successive period of 

 3 years, beginning with 1871, and ending with 1880 ; alsD that 

 the relative speed of the hundred fastest horses in each year is 

 closely the same, though their absolute speed differs. 



We may read the table in another way. If the number of 

 horses that can run a mile in 2 minutes 27 seconds or less is 

 99, we may infer without risk of sensible error that the 99th 

 horse in the order of running accomplishes a mile in that tim, 

 exactly, because the 100th horse certainly takes a longer time, 

 and it is statistically incredible that the rate of the 99th and of 

 the 100th horses should differ by more than a barely perceptible 

 interval. For the same reason we may infer that the 40th horse 

 in that same year runs a mile in 2 minutes 25 seconds, and so on. 

 We can now draw curves, and by graphical interpolation find 

 with the greatest facility the mile rate of the horse in any order 

 of running in any year that we please to select. I have selected 

 the 100th, 50th, 20th, and loth horse respectively for each year 

 beginning with 1874, when we are informed that the returns first 

 begin to be accurate, and have thrown the results into the fol- 

 lowing simple table. The curves obviously required a little 

 smoothing here and there, and in three or four places the read- 

 ings have been thereby modified by one or two tenths of a 

 second. Otherwise they are given directly from the rough 

 plottings. 



Number of Seconds and Tenths of Seconds in Excess of Tuv 

 Minutes that are required for Running One Mile by the 

 Horses whose Order in the Kate of Running in each Year is 

 given at the Top of the Columns 



Mem. — The first horse runs the mile in about 5 or 6 seconds less 

 than the tenth horse. 



It will be found on plotting the figures in the vertical columns 

 into curves, that they run with much regularity and differ 

 little from straight lines. The general conclusion to be de- 

 rived from them is that the improvement of the running 

 shows as yet little tendency to slacken, though no doubt if the 

 number of horses bred for trotting ceased to increase yearly at 

 the same large rate as hitherto, it might do so. Supposing-, 

 however, the conditions to be maintained, I should anticipate 

 that in 1890 there will be about 15 horses that will ran a mile m 

 2 minutes 1 5 seconds or less, and that the fastest horse of that 

 year will run a mile in about 2 minutes 8 seconds. 



Francis Galton 



The Shapes of Leaves 



Mr. Grant Allen's papers in Nature will evidently 

 serve to direct attention to a most interesting subject which 

 hitherto appears to have been much neglected. Every con- 

 tribution of observed facts may tend to throw further light 

 upon it, and I therefore venture to remark that one cause 

 of the frequently filiform character of the leaves of water- 

 plants appears to be the elongating action exercised upon the 

 cells by the pressure of a rapid current of water, since it i» 

 obvious that growth must take place in the direction of the 

 least resistance. With a radiate-veined leaf the tendency niu.-t 

 be towards lateral pressure, which would compress and elongate, 

 and so give a linear form to the leaf-cells. I have been much 

 interested to observe that on the seashore, in places where Fuci 

 are exposed to this action by the ebbing tide, as when growing 

 on the edge of a large boulder or hanging over its sides, the 

 fronds and even the receptacles become unusually elongated. 



