94 



SCIENCE. 



[N. S. Vol. I. No. 4. 



distm-bing influence, cutting off the growth 

 of the trees sharply just before tlie flowers 

 opened. The stately monarchs of the Plio- 

 cene forests began to change their habit and 

 adapt themselves to the new meteorological 

 condition, the ever increasing cold. The 

 unopened flowers were enveloped by the yet 

 undeveloped leaves, which became harder 

 and firmer, forming membranaceous and re- 

 sinous covered bud scales, as a protection 

 against the ice and cold. Flowers thus pro- 

 tected remained dormant during the long 

 glacial winter, and on the return of the next 

 growing season opened their flowers for wind 

 fertilization. This habit of earlj' flowering- 

 became impressed so strongly on the plants 

 that' it became hereditarily fixed. Trees of 

 abnormal habit frequently show atavism, 

 floweriag in the late autumn, if exception- 

 ally warm. This apparently indicates that 

 the cold cut into two periods the normal 

 process of jilant growth. The dJAasion, thus, 

 of the period of growth into two unequal 

 halves by the glacial cold explains why our 

 forest trees have varied little durmg the pro- 

 cess of time fi'om a wind pollinated (anemo- 

 philous) state, because their floral organs 

 are developed in the spring before the ap- 

 pearance of the most highly specialized 

 flower visiting insects. Two causes have 

 operated to keep our trees permanently in 

 an anemophilous condition, first, the sepa- 

 ration of the vegetative and reproductive 

 stages by the cold of the Glacial Epoch, and 

 their early spring flowering ; and secondly, 

 the association of trees together into forests, 

 flower visiting insects loving essentially 

 open glades, or areas devoid of timber. 



More diificulty is experienced in explain- 

 ing the appearance of the herbaceous vernal 

 flora. In order to arrive at a clear under- 

 standing of the problem, a few statistics are 

 necessary. 



The following table compiled from a 

 variety of sources arranged for convenience 

 of presentation according to the system of 



A. L. Jussieu (now little used) will be of 

 use as apjiroximately showing the statistical 

 systematic distribution of our spring jjlants. 



r stamina epigyna 

 Polypetalse \ " hypogyna 



i ' ' perigj'na 



C Corolla hypogyna 

 Monopetaloe j " perigj'na 



i " epigyna 

 (_ Apetalie. 



Jlonocotyledones. 



S e ' ^ . o i 3 



Q P O K,M 



6 1—1 3 

 91 61 86 75 22 



6 18 17 30 16 

 13121 91117 



8 5 913 

 6 15'13 — 



A predominant number of the plants, 

 tabiilated in the foregoing table, fall into 

 eight natural orders: Eanunculaceaj (but- 

 tercup family), Cruciferse (cress familj^) 

 Violacese (violet family), Caryophyllacese 

 (pink family), Eosaceae (rose family), Saxi- 

 fragacete, Ericacese (heath family), Com- 

 positse (sunflower family). The plants 

 belonging to these eight natural orders form 

 the major and characteristic part of our 

 spring flora, and with the exception of the 

 Ericaceae and Comjjositre (few in number) 

 are all polypetalous (many petals, distinct), 

 and monocotyledons hypogynous (stamens 

 and parts below the ovary) in its make-up. 

 The more complex and irregular flowered 

 fandlies appear later in the year. Now this 

 order of flowering corresponds curiously 

 with the order of evolution of the flowering 

 plants, which was suppositiously as follows : 



A. Monocotyledons. Wind Fertilized. Grasses, 

 Sedges. 



B. Dicotyledons. 



1. Wind Fertilized. Ti-ees. 



2. True Insect Fertilized. 



(a) Pol}'pet«ili3e. ( Petals distinct, 4 or 5. ) 



(b) Gamopetalos. ( Petals united. ) 



This comparison leads us to infer the ab- 



1 Darraoh, Proc. Acad. Nat. Sci., Phila., 1860, 145; 

 ''Darlington, Flora Cestrica; 'Gray Manual; ^Roth- 

 rock, Flora of Alaska; ■'Burk, Flora of Greenland, 

 Proc. Acad. Nat. Sci., Phila., 1894. 



