192 THE POPULAR SCIENCE MONTHLY. 



forms, which have each been considered by confirmed "splitters" as 

 distinct species. The first {P. sarmentosa, of Muhlenberg) grows for 

 the most part on very dry soil, and like most plants of arid situations 

 runs largely to pronounced hairiness ; for it is a general rule that 

 water-haunting kinds are smooth and glabrous, while dry or desert 

 types are intensely hirsute (the reason for this wide distinction, 

 though well known, would carry us too far away, this morning, from 

 our main subject). The second form, erected by Michaux into a sepa- 

 rate species (P. simplex) but reduced to subordinate rank as a variety 

 by Torrey and Gray, belongs to moister soil or to deep meadows, 

 where the lush grass prevents evaporation ; and this type grows less 

 hairy and greener, and attains a larger and more luxuriant stature. 

 The two forms differ also in other ways, strictly dependent upon their 

 differences of locality. Sarmentosa^ the dry type, creeps squat 

 upon the ground, as if to avoid the sun. and sends out long, rooting 

 runners in every direction after the fashion of the strawberry-vine ; 

 whereas, simplex, the moister kind, has ascending stems, which rise in 

 competition among the grasses around them, seldom if ever creep, 

 and never produce summer runners. Again, sarmentosa begins to 

 blossom early, and ends early — April to July in the latitude of New 

 York ; while simplex comes and stops a month or so later at either 

 end — May to September in the same district. In other words, the dry- 

 type flowers early in spring on its basking banks, but retires from the 

 scorching heat of your American summer ; while the moist type be- 

 gins later in its shady habitat, but is less affected by the droughts of 

 August. 



Curiously enough, our common European cinque-foil (P. reptans), 

 the exact analogue of your American plant, and fellow-descendant of 

 the self-same pre-glacial ancestor, has also two well-marked forms usu- 

 ally considered as distinct species, but merging into one another by 

 imperceptible gradations. The parent-type (reptans proper) grows in 

 rich pastures or meadows, and answers best to your variety simjjlex, 

 though it sends out long, creeping stems which root every now and 

 again at the nodes ; it has five large petals to each blossom, and the 

 flowers are identical with those of the Canadian cinque-foil. But on 

 open moors, heaths, and dry places, we have a smaller, closer, and more 

 creeping form, the tormentil (P. tormcntilla) ; it is silky-hairy, like 

 your own sarmentosa, and its upper leaves have often only three leaf- 

 lets instead of five, thus reverting to the ancestral type of foliage, 

 when the plant was rather a tre-foil than a cinque-foil. But oddest of 

 all, the small flowers have only four petals, arranged like a Maltese 

 cross ; whereas all their congeners have their full complement of five, 

 in accordance with the old central plan of the entire rose family. Still, 

 the first flower of all on each stem, produced when the plant is in its 

 vigorous youth, has occasionally five petals ; a reversionary fact of 

 great interest. The tormentil has also an intermediate variety of its 



