TRELEASE — OBSERVAT's SUGGESTED BY PRECEd'g PAPER. 29 r 



its flowers, for it is difiicult to say from herbarium specimens 

 whether they are trimorphic, or, as I have thought, merely ex- 

 tremely variable in the relative length of stamens and pistils. 



Measurements to determine the heterogony of a doubtful plant 

 are best made by gathering at random loo or more flowers from 

 as many different plants, and recording the length of each set of 

 stamens and of the pistil of each from the base of the flower, in 

 millimeters and decimals. Projections which show the relations 

 of these graphically, are then readily made on coordinate paper. At 

 the same time, the pollen-grains of as large a number of flowers as 

 possible should be measured : for the general rule is that those from 

 the longest set of stamens are largest, and those from the shortest 

 set, smallest ; and any difference in the size of the stigmas of the 

 different forms of flowers, or of the length of their papillte when 

 they are roughened, should be recorded. In addition to this 

 work, a patient and careful experimenter, who lives where any 

 of our heterogone plants grow abundantly, may do valuable work 

 in determining the relative fertility of what have been called le- 

 gitimate and illegitimate unions between the difl^erent forms, by 

 Mr. Darwin, whose book on different forms of flowers in plants 

 of the same species is the best possible guide for any work of 

 this nature. 



