392 Griggs : Characters of the Plataxaceae 



nodes, and carpels in regular circles, the parts of which alternate 

 with each other. In this figure the divisions of the corolla are 

 about half as long as the carpels — so large that they could 

 scarcely be overlooked on examination even with a hand lens. 

 He supplies diagrams also, showing trimerous, tetramerous, and 

 hexamerous flowers with the parts equidistant in regular circles. 



Only a few stages, however, are favorable for study. He says 

 (p. 309): '^The best time for the study of the completely de- 

 veloped flower is when the anthers have just begun to shed their 

 pollen. Before this period the investigation is enormously diffi- 

 cult ; afterwards the staminate flowers dry up and become unfavor- 

 able for investigation ; likewise the pistillate flowers shed a part 

 of their organs very quickly after anthesis. In fruit a strong 

 growth of hair very similar to the pappus of the Compositae de- 

 velops around the carpels, so that it is no wonder that even emi- 

 nent botanists have stated that the perigynium of the Platanaceae 

 is reduced to hair-like scales between the separate flowers, which 

 represent the floral parts," But even at this favorable period he 

 was able to find no constancy in the regular cycles he figures. 

 On this point he says (p. 311): "In the case cited above where 

 the different organs of the flower occur in like number the con- 

 firmation of the diagrams, empirically obtained, was more or less 

 easy. But on account of the crowding of the flowers all possible 

 cases of displacement, suppression, and stunting of the organs are 

 brought about, so that in many flowers one can determine abso- 

 lutely nothing with certainty regarding the spacial relations of the 

 parts." 



Schoenland did his work only a few years after the appearance 

 of Eichler's Bliitendiagramen, when the influence of Eichler's school 

 was at its height. This school had already pushed the method 

 of studying flowers by means of the comparison of floral diagrams 

 to great extremes. Their method had proved so useful that it 

 was supposed that the flowers of all plants must* be referable in 

 some way or other to such regular schemes. Under the influence 

 of this school Schoenland seems to have labored principally to 

 verify a diagram representing the cycles which he supposed ought 

 to be present. In this deductive method of reasoning he was 

 prevented from checking his results by the facts because of his 



