ARTICLES 



65 



0> 





10 



05 



r- 

 5 



CO 



en 

 5 



o 



<N 



a 



No. of 

 Flowers. 



80 



Let us turn now to the charts of curves. On Chart I are 

 depicted graphically the numbers of plants found in flower 

 on March i — and on this chart we have included the record 

 for 191 1 also and the recent record for 1920. 



The curve speaks 

 for itself : its two 

 outstanding features 

 being the abnormal 

 minimumof 1917 — the 

 result of the bitterly 

 cold winter of 1 9 1 6- 1 7 

 • — and the abnormal 

 maximum of 1920. 



On Chart II we 

 have the curves ' 

 from year to year of 

 the groups A-H, and 

 it seems desirable 

 briefly to call atten- 

 tion to the leading 

 characteristics olf 

 these. Curve A is 

 generally parallel to 

 the March i curve, 

 but inverse in 191 5 

 and 1916 — thus indi- 

 cating that, although 



1915 was behind 

 1914 on March i 

 {i.e. in February and 

 January), it was more 

 forward by the end 

 of March ; and that 



19 1 6 was more for- 

 ward than 191 5 and 

 1914 up to March i. 



76 



72 



68 



64 



60 



63 



62 



43 



44 



40 



36 



32 



28 



24 



Chart i. 



but behind both by the end of March. 



1 These were drawn on the vertical scale of i centimetre : 2 days. 



To economise space the curves have been telescoped into one another 

 as far as possible, so that, e.g., day 44 on A and B and day 40 on C are at the 

 same level ; and in all 165 days have been telescoped into the space that 

 would otherwise have accommodated only 83 ; but in each curve the scale 

 of one centimetre : 2 days has been preserved. To give one illustration — 

 curve A ranges from days lo-ii (March 10 and 11) in 1918 to between 

 days 45 and 46 (April 14 and 15) in 1917. 



The process of averaging necessarily yielded in various cases such results 

 as 1 03 for a date. 



The alternative dates on A for 19 18 and B for 19 14 — giving rise to the 

 alternative dotted curves — indicate the inclusion or exclusion of the species 

 specially marked ^ and ^ in the tables of average dates. 



5 



