[Vol. 9 



248 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



its highest development in C. lasiophyllus, but even here it is 

 trivial and serves rather for confirmation of relationship than as 

 a definite diagnostic character. The small stigmas that termi- 

 nate the styles of this type serve to emphasize this character. 



The degree of lobing of the stigma is of considerable diagnostic 

 value. Except in those species with conic styles, in which it is 

 uniformly and minutely two-lobed, the stigma varies from en- 

 tire to deeply two-lobed. In these species the lobing is in- 

 variably such that the lobes are produced over the valves. The 

 lobes are always much more evident in the mature fruit than in 

 the flower. This is perhaps due to the collapse of parenchymous 

 tissue on drying and the consequent emphasizing of the vascular 

 strands of the style. In the descriptions the length of the lobes 

 of the stigma is derived from measurements in the fruit. It 

 is believed that the entire stigma represents the primitive type 

 and the degree of lobing is a measure of specialization. In Thely- 

 podium, which is to be considered more primitive than Caulan- 

 thus, the stigma is almost universally entire. Those species of 

 Caulanthvs that are evidently aberrant and specialized— as C. 

 heterophyl tus and C. calif amicus— exhibit a definitely lobed 

 stigma. 



Septum.— The cell pattern of the septum in Caulanthus is 

 quite uniform except for two or three species. It is thin and of 

 short inost.lv straiirht-walled cells of ncarlv equal breadth and 



length. This is in striking 



f Th el Imodium 



which there is a highly differentiated middle band. The excep- 

 tions referred to occur in C. anceps and C. lasiophyllus. In 



the first species the cells are not unlike the normal type for the 

 genus except that the walls are somewhat tortuous rather than 

 straight. In the second the walls are tortuous and the middle re- 

 gion is differentiated as in Thelypodium. If one were to judge 

 from the septum alone C. lasiophyllus would have to be retained 

 in Thelypodium where it has previously been carried. 



Seeds. — Few characters of taxonomic value are found in the 

 seeds. In most of the species they are neither winged nor mar- 

 gined, and usually the cotyledons are somewhat oblique with 

 respect to the radicle. This is the condition found 

 dium exactly. The relative position of the 



cotyledons and 



is often quite variable. In C. pilosus, for instance, the pos 

 may vary greatly in seeds of the same individual. The most 



snifMinns pyp.entions to the usual structure are seen in C. he 



