So 



KNOWLEDGE cS: SCIENTIFIC NEWS, 



0~ 



[ArRii., 1905. 



An interesting question remains as to what becomes 

 of this fluid, for at a later stage the wings are per- 

 fectly dry. Some of it probably escapes through the 

 pores of the wing, but I am inclined to think that the 

 greater part is withdrawn into the abdomen by an 

 action the reverse of that by which it was injected. 



For hours after the wings are perfectly flat and dry 

 the abdomen is greatly distended throughout its whole 

 length, and at intervals during the first day after 

 emergence, fluid is ejected, a drop or two at a time, 

 from the rectum. This fluid is perfectly colourless, the 

 green colouring matter being retained in the body of 

 the insect (in this case of ^Eschna Cyanea or green 

 insect). And not for several days does the abdomen 

 become dry and quill-like as when we see the insect 

 on the wing. 



Forecasting Seasons. 



To THE Editors of " Knowledge." 

 Sirs, — While there are many weather prophets, we seem to 

 have at present little or no sound knowledge as to the character 

 of future seasons, though the cyclical nature of much of our 

 weather, on which fresh light is being thrown, gives reason to 

 hope that this important art of long-range forecasting will one 

 day be achieved, in some measure. 



I have lately met with a case in which, I think, one could 

 feel something like certainty regarding a forecast of distant 

 weather ; and a forecast sufficiently definite to be useful. 

 The subject is that of the number of frost days at Greenwich 

 in the latter half of 1904 (the average in that half being 

 about 18). 



J ISvi '6 'so 'h 'g ^^•^ '6 '7° '•' 'S 'y^ '6 ''?'» '^ 9 oa fiPl, 



For such inquiries I have been making use o( lii'iccsmoolliid 

 curves. Thus in the present case, the series of numbers of 

 frost days in the latter half, from 1841, is first smoothed in 

 sums of five (grouping 1841-45, '42-'46, and so on). Then 

 these sums are smoothed in the same way. The sum in each 

 case is put opposite the middle member of the group. 



The series thus obtained yields the lower curve of the 

 diagram. The upper curve is got in the same way from the 

 numbers of frost days in the earlier half of the year (average 37). 

 Above are indicated the years of maximum and minimum sun- 

 spots (without regard to numerical relations), the influence 

 of which, I believe, comes out in these curves ; but without 

 entering into this question, or considering how this double 

 smoothing affects the truth of Nature, I would merely call 

 attention to the continuous up and down course of the curves — 

 the rise for several years from the lowest points, and fall for 

 several years from the highest. 



The lower curve, ending in 1899, is derived from the actual 

 series ending in 1903. Now, in prospect of the latter half of 

 19<J4, we might ask, What will the curve do next, go up or 

 down ? Few. probably, would hesitate to say, go up. Then 

 how much will it go up ? Here it often seems difficult to form 



a right conjecture. But in the present case it so happens that 

 a rise of only i, that is, an addition of one day to the figure for 

 iSgg, means that the actual number of frost days for 1904 

 would be in excess of the average (over 

 figures from 1S90 to make this clear : 



1892 



94 



1S93 

 14 

 75 



1901 



1894 

 II 

 79 

 374 

 1902 



1S95 



14 



65 

 348 

 1933 



1 will give the 

 1896 



iS 



61 

 330 

 1904 



(20) 



1S97 

 8 

 68 

 314 



1900 



3 24 10 



65 67 (6g) 



()20 

 (a) Actual numbers. (/') First smoothing, (c) Se-ond smoathing. 

 The figures in brackets are those for an addition of i to the 

 curve-value of uSgg, giving the value 20 for (ii) in 1904; 2 above 

 average. 

 The actual number proves to be 23 ; 5 above average. 

 One could thus safely predict a cold last quarter (frost days 

 over average), and might even estimate the amount of excess 

 approximately. 



Doubtless other cases equally clear may be met with. 

 I am, yours, &c., 



Alex. B. MacDowall. 

 g, Saltwood Gardens, Ilythe. 



Wireless Telegraphy. 



Those particulars that have been made public of the opera- 

 lions of wireless telegraphy in the Russo-Japanese war have 

 amply served to endorse the soundness of the official view 

 that all the wireless tele{;raphic installations of a country 

 should be licensed and known, so that in case of war they 

 could be immediately brought under Governmental supervision 

 and control. We are in a position to say that the Government 

 Wireless Telegraphy Bill, which was at first vigorously con- 

 demned by more than one electrical engineer as likely to 

 hamper and stiHe enttrprise, has been found in practice to do 

 nothing of the kind, and is now acclaimed by some who were 

 at first its opponents. Tlie Act is being administered in a 

 broad-minded spirit, and it has been recognised that it 

 operates for the convenience of investigators. " Before the 

 Act," Professor J. A. Fleming has recently said, " we were in 

 the position of a number of people at a public meeting who 

 might by all speaking at once prevent anyone from being 

 heard. The Act regulates and distinguishes our utterances," 

 which is another way of saying that as yet wireless telegraphy 

 has not yet reached the stage when it is secure against violent 

 " interference " from conflicting stations. The same is true 

 in a less degree of ordinary telegraphy along wires, which 

 might be upset by anyone rich enough or malicious enough to 

 set the requisite quantity of electric energy in motion. But 

 it is the fact that wireless messages, if regulated and controlled, 

 can be sent and received without interfering with other wire- 

 less stations with which they are not concerned. The next 

 step in wireless telegraphy will be to ascertain the jioint of 

 origin from which a wireless wave is sent. That can be done 

 to a limited extent now, but it will become more easily accom- 

 plished when the measurements of electric waves and the 

 measurements of the sources of energy in, and produced by, 

 wireless telegraphic instruments can be much more accurately 

 measured. Tho present stage of wireless telegraphy resembles 

 that of cable telegraphy before electricians like Lord Kelvin 

 and Mr. Latimer Clark had shown that accuracy and distance 

 could only be attained by instruments of measured refine- 

 ment, and that the first step to this daidcratum was refined 

 measurement. 



Star Maps. 



In our next issue (May) we propose to coninu'nce the 

 series of new star maps. These will be on the system 

 used in Proctor's Atlas, comprising in all twelve maps. 

 The stars will be in white on a dark blue ground so as 

 to stand out clearly, oM-r which will be printed the 

 names of the constellations, letters (and some names) 

 of stars, and R.A. and Declination. 



