438 KINSHIP AND ADAPTATION 



carpels, numerous stamens of ordinary form, and a regular perianth 

 of almost leaf -like sepals; and a very simple form of palmate leaf 

 and herbaceous stem. Certain other descendants of the primitive 

 stock showed in very early times a modification of the torus and 

 perianth (resulting in pcrigynous flowers with the perianth differ- 

 entiated into a well-defined calyx and corolla), also an advance in 

 complexity of leaf-form, and more or less woodiness of stem. The 

 plants of this branch were thus able to thrive imder the more exact- 

 ing conditions of open fields, and have become the highly developed 

 sturdy peonies of to-day. Very early in the development of what 

 we may call the hj^pogynous branch of the family, there appeared 

 some more or less Caltha-like forms, in "which the carpels did not 

 open at maturity, but separated like seeds from the parent plant. 

 This change of habit made it unnecessary for more than one seed 

 to mature in each carpel, so the result was an achenial instead of a 

 follicular fruit. Those descendants which retained rudiments of 

 several ovules became in time either forms of Anemone or of Clem- 

 atis according as they developed the peculiarities of this or that 

 genus; while similarly those in which the reduction to one ovule was 

 complete, gave rise to such plants as buttercups and mouse-tails. 

 In much the same way we find those descendants which comjirise 

 the more primitive subbranch with its many-seeded carpels becom- 

 ing differentiated into forms retaining the original follicular fruit, 

 and forms in which the fruit is indehiscent and fleshj^, as in Aetata. 

 Along the lines with follicular fruit the appearance of forms with 

 certain of the stamens changed into more or less conspicuous nec- 

 taries give rise to a departure from the Caltha-like forms, which in 

 turn became differentiated into those in which the corolla remained 

 regular and those in which more or less irregularity v/as shown, as 

 in monkshood. Finallj^, the regular-flowered forms with staminodes 

 developed such peculiarities as the spurs of Aquilegia, and the 

 curious follicular capsule of Nigella, bj' which the modern genera of 

 this subgroup are now distinguished. ^ 



1 It might fairly be asked whether entirely different lines of descent 

 from those here given would not as well explain the modern forms upon 

 which our reasonings have been based. Undoubtedly this is true. Thus, 

 instead of supposing the progenitor of the family to have been an herb 

 like tlie marsh-marigold we might perhaps with more probability assume 

 it to have been a shrub or tree resembling a woody peony or a magnolia; 

 for in other families there is much evidence in favor of the view that 

 herbaceous seed-plants have had woody ancestors. But in this family 

 we have no direct proof that its earliest members were wood}', and for 

 us to make the assumption would complicate our reasoning without 

 rendering any clearer the general principle we are trying to make plain. 

 Let the student, therefore, take our crowfoot family tree simplj' as rep- 

 resenting one out of many possible ways of accounting for the facts at 

 hand, and as liable to modification whenever new light appears on the 

 problem. It is merely intended to illustrate the kind of reasoning that 

 biologists emjiloy in the absence of evidence from fossil r(>mains, and 



