64 SCIENCE PROGRESS 



are excluded — although the totals of these are dealt with 

 separately. Others were rejected because they can be found 

 only at a distance which involves a special expedition, and 

 notoriously are not usually noted until they have been in 

 flower for some time ; whilst others again were rejected because 

 one or two records were so aberrant in date as to suggest 

 that in one case there had been an abnormally early " sport,", 

 and in another some accidental delay in looking for the flower. 

 After all such rejections some ninety species were chosen 

 and arranged in seasonal groups ; but, since the list still included 

 some species with an abnormally large time-range, these were 

 critically discussed by us, and several rejected as unreliable ; 

 and finally, after the average time-range of each group had been 

 determined, any species whose time-range exceeded this average 

 by 50 per cent, was struck out. The curves ultimately obtained 

 indicate very satisfactorily the reliability of ourthus sifted data. 

 In the preceding tables we exhibit the seasonal groups, 

 the time-range in days of each species and its average date of 

 flowering during the eight years 191 2-19, the average time-range 

 of the group, and the average date of flowering— reckoned in 

 days after February 28 or 29 — of the group in each year. 

 These average group-dates are the data from which the curves 

 are constructed. Prefixed to the name of each species also is 

 the number of records thereof during the eight years ; for 

 owing to occasional absences from home some species were 

 sometimes collected elsewhere ; and in such cases there are 

 necessarily blanks in the phenological record for this place. ^ 



^ As regards species found in flower on March i, it must be understood 

 that these fall into several distinct series, viz. : 



(a) Species always or normally in flower at this date here. 



{b) Two or three species occasionally found in flower in exceptionally 

 forward seasons. 



(c) Stragglers, or survivors through a mild winter from the previous 

 year. Heracleum spondyllium and Lychnis dioica are characteristic normal 

 survivors ; but, since some species are found as survivors in one year and 

 others in another, the total number thus unavailable for phenological pur- 

 poses — for no species has been selected that is sometimes found as a survivor, 

 since no record is made in the Wild Flower Society diary of the first flower 

 (in the new season) of such species — is appreciable. 



[d) Occasional obvious sports; e.g. the Ox-eye Daisy has been found on 

 March i. 



Now, (a) and (c) are necessarily excluded from the groups. As regards 

 (d), in perhaps one or two cases the species has been included for the years of 

 normal flowering, and the " sporting " year treated as a blank. The only 

 real perplexity has been caused by (6) ; and after consideration these species 

 have been included in the groups as well as in the March i tables — but 

 in each case a note has been appended to the table. The fact is that, after 

 the exclusion of species normally or frequently found on March i, so few 

 March flowers were left that we could not afiord to cut out these ; but in 

 at least one instance — as we shall have occasion to point out — such inclusion 

 of one species appreciably affects a curve. 



