36 



brings up the following problem : Was the primitive form from 

 which the modern Linaria evolved one with five spurs, or did the 

 normally single spur evolve after the onset of zygomorphy? The 

 present writers are inclined toward the latter viewpoint.^ 



Other flowers on the Handley specimen (as Figure 5 with its 

 bifurcate spur and Figures 6a-8c with supernumerary anthers 

 and incomplete peloria) are of interest if only to indicate the poten- 

 tiality for plasticity of ontogenetic variation within the developing 

 organs of a single individual. It will be noted (particularly in 

 Figures 6a-8c) that there seems to be a spatial relationship between 

 the production of functional supernumerary anthers and the devel- 

 opment of supernumerary spurs. The external views and floral 

 diagrams of these flowers will be self-explanatory. 



During our examination of herbarium material we came upon 

 a series of specimens collected by R. Hitchcock near Van Cort- 

 landt Park, New York City, in June, 1909. Although no two 

 flowers of this collection were exactly alike, it was unusual in that 

 the various ones examined had the same general types of abnormali- 

 ties : they were smaller than is normal for Linaria vulgaris ; the 

 spur was lacking or but poorly developed ; the corolla, although 

 bilabiate, was irregularly erose; and, still more remarkable, in 

 addition to the four normal stamens, additional rudimentary stamens 

 or anther pouches were found (see Figures 12a-14e). In every 

 case these were attached to or partly imbedded in the tissues of 

 what ordinarily would have been the lower of the two corolla lobes 

 {i.e., the one composed of three corolla segments, the middle one 

 of which ordinarily would have borne the spur). 



There is but little point in discussing this phenomenon with- 

 out having worked on the living material, but studies some years 

 ago by the senior author on other species (principally Cannabis 

 sativa, the results of which are not yet completely published) indi- 

 cated a close relationship between a disturbance of the normal 

 sequence of metabolic rates within developing flower primordia 

 and the production of such monstrosities. It is to be here noted as 



^ The reader will undoubtedly think of analogous problems, as for example 

 in the Ranunculaceae where Delphinium has one of the perianth segments 

 produced into a spur, or in Aqiiilcgia where spurs are found on five perianth 

 segments (although admittedly in a different cycle from that of Delphinium) . 

 The genus Aconitum is also subject to peloric forms. 



