88 PLANT MORPHOLOGY 



chalazogamy, which elsewhere occurs in forms which may be re- 

 garded as degenerate; the facts that only a few of the ovules develop 

 further; that at the time of anthesis they are in many forms not 

 yet present; and finally the dicliny of the flowers. There has been 

 much contention over the question whether the androgynous flowers 

 of these forms are to be admitted to be the original form or not. 

 Let us look at, e. g., the Cupuliferce. Most of the forms have dicli- 

 nous flowers. In Castanea vesca, however, androgynous flowers 

 occur regularly: in the male flowers rudiments of the ovary, while 

 in the female flowers staminodia, are often evident. But we know 

 that for reduced organs all gradations occur from nearly complete 

 development to almost entire disappearance. From the formal 

 standpoint, then, the androgynous flowers may, with at least as 

 much justice, be regarded as primitive as the diclinous ones, which 

 more recently have been thus branded. 1 Just this question is, 

 however, fitted to clear up the difference between pure phylogenetic 

 and causal morphology. The latter says: By the mere comparison 

 of forms morphological questions may not at all be decided. We 

 must first of all become more closely acquainted with the forms to 

 be compared, by seeking to determine the conditions under which, 

 in living plants, the configuration of parts is produced. Concerning 

 the flowers of the Cupuliferce the question then arises: is the occur- 

 rence of male and female flowers dependent upon different conditions, 

 and are these other than those under which androgynous flowers arise? 

 As a matter of fact, it may be determined that, e. g., in the oak, the 

 female flowers always occur in those parts of the twig which are 

 stronger, that is, better nourished, than those in which the male 

 flowers occur. This offers us, however, only a point of departure for 

 a more exhaustive research. When we know better the relation 

 between the formation of flowers and the total activity of the plant, 

 when we have the ability at will to cause it to produce male, female, 

 or androgynous flowers, when we further know how it is determined 

 that the oak usually brings to development only one out of six 

 ovules, and why the pollen tube follows a different path than the 

 usual, then may we further discuss the question whether the Cupu- 

 liferce are primitive or not for then shall we have better grounds 

 for phylogenetic conclusions than we have at present, and we shall 

 then recognize with great probability the changes which have taken 

 place in these organs as phenomena resulting from changes in the 

 total organization of these plants. 



1 It was chiefly the fact that among the Gymnospermce the flowers are typically 

 diclinous that led to the view that this condition is the primitive one for the lower 

 AngioKperma '. In making such comparisons, however, we should always start 

 from the group in question, and not from some other one. In a very great number 

 of cases dioliny is certainly not primitive in angiospermous flowers, so that it ap- 

 pears entirely probable that also in the remaining cases dicliny is to be derived from 

 raonocliny. 



