Vol. 5 ^o. 12 



TORREYA ..BRA«v 



NEW YORK 



December, 1905 



rtOTANICAL 

 GARDEN. 



A STATISTICAL METHOD I-OR COMPARING 11 11-: 

 AGE OE DIEEICRI'.NT E LOR AS 



l>v Rui.ANi) M. IIakpkr 



It is a well-known principle of ])hytogeography that when an 

 area devoid of vegetation and true soil, such as one which has 

 recently been covered with water or ice for a long period, is first 

 invaded by plants, the lower forms tend to predominate at first, 

 and gradually pave the way for higher ones.* It is also gen- 

 erally conceded by botanists that monocotyledonous plants as a 

 class are of lower rank than dicotyledons. Putting these two 

 conceptions together, a method is at once suggested for deter- 

 mining roughl)' the age of a given flora, for a study of the rela- 

 tive proportion of monocotyledons and dicotyledons in any essen- 

 tially homogeneous region ought to throw some light on the 

 length of time that that region has been continuously occupied 

 by vegetation.! The application of this method, crude as it may 

 seem, gives some remarkably consistent results for regions be- 

 lieved to be of the same age geologicallx'. 



The glaciated region of the northern states is believed to have 

 been entirely devoid of vegetation — at least as far as flowering 

 plants are concerned — as late as fifteen or twenty thousand years 

 ago ; and most of the coastal plain of the southeastern states was 

 probably submerged beneath the sea at about the same time. 



* Prof. N. S. Shaler's very interesting paper on "The origin and nature of soils" 

 {i2th .Ann. Rep. U. S. Geol. Surv., pp. 213 et seq. ) should be consulted in this con- 

 nection. 



fThis method is so simple that it can hardly be claimed as original, but it prob- 

 ably has not been applied to so many different parts of Eastern North America before. 

 MacMillan came very ne^r it in some of the statistical discussions in his " Mela- 

 spermae of the Minnesota Valley" in 1892, but did not use it for comparison in this 

 way. 



CO [No. II, \'ol. 5 of TORKliVA, comprising pages 1S7-206, was issued Novem- 



g ber 25, 1905.] 



"- 207 



