22 



THE AMERICAN NATURALIST 



[Vol.XLIV 



would be expected in accordance with some mathematical 

 series. It is obvious to any one trained in working with 

 numbers that there are serious difficulties in the way of 

 demonstrating a number of modes in a frequency distri- 

 bution. The errors of random sampling can not be dis- 

 regarded and when the variation is continuous instead of 

 discrete the chances of errors due to biased judgment 

 are great. These facts emphasize the necessity for 

 seeking out the simplest possible cases of polymorphism 

 and determining in so far as possible the morphological 

 conditions which underlie them. Such a case seems to be 

 offered by the involucral leaves of Syndesmon. 



After I had made my first series of countings a paper 

 by Kellerman 14 giving the essential morphological fea- 

 tures of the leaves and considerable statistical data con- 

 cerning their variation came to my notice. He does not 

 present his data in a form suitable for statistical con- 

 sideration, and did not note the bimodal condition which 

 appears where a curve is plotted for the entire number 

 of lamina?. 



K el Ionium's diagrams represent the morphology of 

 the inflorescence very well and are reproduced here with 

 such additional ones as are necessary to represent the 

 types of leaves observed in our series. The terminal and 

 axillary buds which may develop into flower-bearing axes 

 are represented by solid dots. The number of leaflets 

 into which a leaf is divided is indicated by the partial 

 division of the line representing the leaf. Unfortunately 

 Kellerman 's data are rather too few to be given further 

 statistical analysis. For figures illustrating the general 

 appearance of different inflorescences the reader may 

 consult his plate. 



Most simply the inflorescence consists of a terminal 

 flower, two involucral leaves and two axillary flowers. 

 The axillary buds do not always develop. The leaves 

 may be entire or divided into (generally) three leaflets. 



The form of inflorescences with two leaves, one ter- 



14 Kellernian, W. A., Ohio Nat., Vol. I, pp. 107-110, 1901. 



