210 



NATURE 



[February i6, 1922 



Flowering Dates of Trees along Main British Railway Routes. 

 By J. Edmund Clark. 



A FEW months ago the Editor of Nature sent 

 me, as joint editor of the Phenological 

 Reports published by the Royal Meteorological 

 Society, an interesting problem. He said: — 

 " It has been pointed out on several occasions 

 that in travelling from the West of England 

 (Devon and Cornwall) to London, fruit trees are 

 usually seen to be much more forward as regards 

 flowering nearer London than in the West." I 

 was therefore asked whether I had, among the 

 " phenological observations, records of such 

 flowering dates arranged according to, longitude 

 so as to determine whether fruit trees do, 

 normally, bloom earlier near London than far 

 away, and whether this is true also in passing 

 from London to the East." 



Our thirty years' records give no basis for a 

 direct reply, since of set purpose garden flowers 

 and fruits" were excluded by the late Edward 

 Mawley from the thirteen selected blossomings. 

 These begin with the hazel (mean for the British 

 Isles, February 13) and close with the ivy (Octo- 

 ber 2). The many varieties of most of our fruit 

 trees is the obvious ground for their exclusion. 



It seemed, however, worth attacking the 

 problem indirectly, even though at first the sup- 

 posed earliness appeared to be improbable. The 

 four fruit-time plants in our list, blackthorn, 

 garlick hedge-mustard, horse chestnut, and haw- 

 thorn, average two and a half days earlier in 

 England, S.W., than in England, S.E. and E., in 

 our thirty years' means. 



The basis for investigation required : — 



(i) Sufficient stations. For the first time in the 

 thirty years 1920 supplies these, thirty-five being 

 available. 1 



(2) The blackthorn gives us the opening, the 

 other two trees the closing, stages of fruit 

 flowering. 



(3) Unfortunately 1920 was abnormally early, 

 and therefore prolonged. This is shown by the 

 following table, giving the days early compared 

 with the mean : — 



1920 S.W. S.E. E. 



Blackthorn, early . . . . days 30 24 20 



Chestnut and may, early . . ,, \o\ 18 12 

 Prolongation, fruit flowering . . ,, igj 6 8 



The possible results on relative conditions may 

 well be serious, especially as to insolation. 

 Obviously any such effect would be most marked 

 in England, S.W. 



(4) The isophenes (lines of equal flowering date) 

 have for the first time been tested for agreement 

 with Prof. Hopkins's Bioclimatic Law, recently 

 formulated and found to hold well in the United 

 States. Starting- from a given station, this postu- 

 lates a retardation of four days in flowering for 

 every additional 400 ft. of altitude, 1° of latitude 



1 See our Phenological Report to the Royal Meteorological Society for 

 1920 (Part 3), 1921. 



NO. 2729, VOL. 109] 



and 5° eastward in longitude^ ; the reverse for nega- 

 tive values. Both he and we have found it fairly 

 trustworthy for the British Isles, and he for 

 Western Europe. This has enabled me to reduce 

 the records to sea-level, so as to see whether there 

 is any factor other than the higher elevations along 

 most of the way until London is approached. 



(5) Since it is most difficult faithfully to record 

 the average date of first flowering, after taking 

 the mean of the three trees for each of the thirty- 

 five stations available (those within twelve to 

 fifteen miles of the selected railway routes), the 

 mean of two or more adjacent stations was taken, 

 so far as possible. In this way we have fifteen 

 sets of records available. 



(6) Unfortunately the L.S.W.R. route from 

 Exeter to London was useless, no records being 

 available between Exeter and Fleet. 



The results are shown in the following table : — 



Mean Flowering Bate, Sloe, Chestnut, May, 1920. 



Stations Date 



I. Penzance, Camborne, Falmouth . . April 8 



II. Polperro (2), Duloe . . . . . . ,, 6 



III. Launceston, Hexworthy, Tavistock . . ,, 6 



IV. Tiverton, Wellington, Taunton . . . . March 30 

 V. Winscombe, Portishead . . . . . . April 2 



VI. Falfield, Bath I 



VII. Winsley, Marlborough . . . . . . ,, 6 



VIII. Oxford (2) „ 3 



IX. Fleet, Farnham, Hampton Wick . . March 29 



X. A. Harrow, Watford . . . . . . April 9 



B. Purley (Surrey) (2) . . . . . . „ 3 



XI. A. New Barnet, Woodford . . . . ,, 2 



B. Hayes and Bromley (Kent) . . . . March 28 



XII. A. Maldon, Hatfield Peverel, Lexden . . April 5 

 B. Maidstone . . . . . . . . ,, 2 



I. to IX. closely follow the G.W.R. from Penzance to 

 London ; X.a, XI.a, and XII. a continue on north of the 

 Thames Estuary to Colchester ; X.b, XI. b, and XII. b 

 south of it to Maidstonci 



Of I. to IX. the earliest date is just before 

 reaching London ; the latest in the foot of Corn- 

 wall. East Devon and Somerset fall little behind 

 London ; whilst East Cornwall, West Devon, and 

 North Wilts are nearly as late as I. Hopkins's lati- 

 tude and longitude corrections, however, intensify 

 the contrast, the date relative to London working out 

 to April 18. 



Eastwards from London flowering dates north 

 of the Thames are retarded, but to the south 

 scarcely at all, Hayes and Bromley (Kent) — 

 March 28 — having the earliest date of all. 



Before considering possible explanations, two 

 cautions must be reiterated. The transfer here of 

 Hopkins's Law values is still only tentative, and 

 the dates of observation may well be subject to a 

 margin of error of a day or two. Making full 

 allowance for these, we may still believe that 

 general observation is verified, the more so that, 



• As a fact, the law_ is more general, governing all seasonal plant 

 phenomena ; the regressive phases, of course require the reverse of the 

 ahove statement. 



