242 



KNOWLEDGE. 



[October 1, 1897. 



days of northerly wind in the first quarter (the days of 

 wind being classified according to the four cardinal points), 

 and smooth it with averages of five, we have the curve 

 marked c, which has a generally similar course to the 

 smoothed curve of frost days. 



Now, the method of smoothing curves, which is largely 

 resorted to in meteorology, appears to be regarded sus- 

 piciously by some minds, as a sort of " cooking " of 

 statistics. Let us then put aside these curves for a little, 

 and take a difierent point of view. 



Consider the following five-year groups: the year just 

 before each minimum of sunspots, and two years on either 

 side. Were the first quarters in those years severe ( + ) 

 or mild (— ), (referring to an average) '? The following 

 table supplies an answer : — * 



Alg-ebraic sum of 

 Min. differences- 



1841-44 ++ — + ...+ 10 



1853-57 -f — -f -- + ... -f 27 



1864-68 + -\- — + — ...+ 7 



1875-79 + + — — + ■•• +4 



1886-90 + + + + — ■■■+ 52 



Thus we find sixteen of those quarters severe, and only 

 eight "'//(/ .- also a preponderance of plus signs in each five- 

 year group, and an excess in each of the algebraic sums. 



Consider next these five-year groups: the year just 

 after each maximum of sunspots, and two years on either 

 side. What kind of first quarters occurred in those years '^ 



Ali.'f'iraJo sum of 

 Mix. Difi'erences. 



1847-51 + + — + — •■■ — 4 



1859-63 — -f — — — ••• — 52 



1869-73 — -i- — — — •- — 28 



1882-86 — — — av. + ... — 26 



1892-96 -t- — — + — ... — 6 



In this case we have sixteen of those first quarters mild, 



eight severe, and one average ; further, a preponderance 



of minus signs in the five-year groups except one, and a 



deficit in each of the algebraic sums. 



The reader is invited to study these facts and diagrams 

 independently, and to estimate the probabilities involved. 



It seems to the present writer that we are now about 

 the beginning of another of those long waves of the 

 smoothed curve of frost days, which may be expected to 

 culminate near the next sunspot minimum. At what date 

 may we look for that minimum ? Taking the eleven 

 years' period as our guide, about 1900 ; but the interval, 

 it muse be remembered, varies. 



We might fairly, then, perhaps expect in the five-year 

 group centred about that time "(if a minimum), or rather 

 before, a recurrence of the state of things indicated in the 

 first of the above tables ; speaking generally, a preponder- 

 ance of cold (as measured by the number of frost days). 

 And lookicg still further to the next maximum of sunspots, 

 if that occurred about 1901, we should anticipate then 

 a preponderance of mild weather, as in the second table. 



As to the degree of excessive cold about this next sun- 

 spot minimum, one may hesitate to give an opinion ; but 

 tlie new wave seems at least hkely to be lower than the 

 last, and more of the order of the two previous. (In this 

 connection it is instructive to consider the thirty- live years' 

 period of weather demonstrated by Brueckner.) 



Should these speculations appear to be too indefinite to 

 be of much practical use, they may yet perhaps incite 

 some readers to further inquiry into this interesting subject. 



* The i-nns-pot rainimum years aie 1843, 1856, )&67, )S78, 1889; 

 the maximum vears, 1818, 1860, 1870, 1883, 1893. In the first group 

 of the first table the value for 1840 is not given, as the scries 

 commences with 1841 . 



ON THE VEGETATION AND SOME OF THE 

 VEGETABLE PRODUCTIONS OF AUSTRAL- 

 ASIA.-IV. 



By W. BoTTiNG Hemsley, F.R.S., F.L.S. 

 Illustrated by Miss Ethel W.\lker. 



IN my last article I announced my intention of devoting 

 my next to the singular and highly diversified family 

 of plants known to botanists as the Proteacece. It 

 is difficult to convey an idea of the aspect and 

 peculiarities of this singular family, because familiar 

 subjects for comparison are wanting, and I can only again 

 recommend those persons in whom my articles may 

 perchance have awakened some interest, to visit the tem- 

 perate house, the Marianne North Gallery, and the 

 museums at Kew, or 

 the botanical depart- 

 ment of the Natural 

 History Museum at 

 South Kensington, 

 where living plants, 

 paintings, or dried 

 specimens are ex- 

 hibited with more or 

 less descriptive and 

 explanatory labels ; 

 but, as comparatively 

 few persons have op- 

 portunities of visiting 

 these admirable and 

 instructive national 

 establishments, I will 

 endeavour to impart 

 some information 

 with my pen, aided 

 by a few figures. 



The P/otfac('<j, with 

 very few exceptions, 

 are confined to the 

 southern hemisphere, 

 inhabiting Austral- 

 asia, Africa, and 

 America, but not ex- 

 tending to Polynesia 

 proper. Altogether 

 about a thousand 

 species are known, 

 and they are referred 

 to fifty different 

 genera ; considerably more than half of them being 

 restricted to Australia proper. Two are found in New 

 Zealand ; between thirty and forty in New Caledonia ; 

 about half a dozen in New Guinea; and one genus 

 extends northward to South China and .Japan. There 

 is also a great concentration of I'roteacece in South 

 Africa, among them the famous silver tree, LeurinhiKlrcn 

 iirgenteiim, a native of the Table Mountain ; and a few 

 stragglers reach as far north as the mountains of Abyssinia 

 and Guinea. In America they are much less numerous, 

 but there also a few species occur north of the Equator, as 

 far as Central Mexico. The genera of each of the three 

 areas — Australia, Africa, and America — are nearly all 

 different, and all the species are diii'erent. Such is the 

 present distribution of the Proteacem, and there is no 

 convincing evidence that any member of this family 

 inhabited the nortliern hemisphere in earlier geologic 



Fig. 1. — A Cone of a S]iecies of Banksia, 

 on whieh only a very few of tlie Flowers 

 have given biitli to Setd-vessels. Nearly 

 natural size. 



