IMPARIPINNATE LEAVES IN CASSIA. 1 39 



APPARENTLY IMPARIPINNATE LEAVES IN CASSIA. 



By J. Arthur Harris, 

 Missouri Botanical Garden. 



As Penzig has pointed out in the general introductory remarks 

 to the Leguminos?e in his Pflanzen-Teratologie, the compound 

 leaf of this family is subject to many anomalies. Trifoliate 

 leaves are characteristic of whole groups, while in others only 

 those pinnately divided are to be observed. It is interesting to 

 find that transitions between these types sometimes occur by the 

 division of some of the leaflets and the elongation of the rhachis. 

 If the terminal leaflet of an imparipinnate leaf be the one divided, 

 an apparently paripinnate leaf may result. Such have been fre- 

 quently observed for Amorpha, Robiuia (locust), Caragana, 

 Sophora and Glcdiischia (honey-locust). The lateral leaflets are 

 also often divided, as in Robinia, Astragalus, Coroiiilla, Glcdit- 

 scli'ia and others. Transitions between pinnate and bipinnate 

 leaves are occasionally observed. Simple leaves are sometimes 

 seen in species normally with compound leaves. Rarely the ter- 

 minal tendril of such forms as Lathyrus and Pisiiiii is found 

 transformed into a typical leaflet or into an ascidium. The 

 development of tendrils in the place of lateral leaflets is also ob- 

 served in Vicia and Lathyrus. Foliar and even stipular ascidia 

 are observed. ' 



During the summer of 1904 I observed on plants of Cassia 

 Sophcra, cultivated at the Missouri Botanical Garden, a few per- 

 fectly formed imparipinnate leaves. As is well known, the leaf 

 of Cassia is regularly paripinnate. The rhachis, however, pro- 

 jects slightly beyond the insertion of the last pair of leaflets in 

 a small subulate or filiform process which might appear to the 

 observer to be homologous with a terminal leaflet. 



The first supposition in regard to these terminal leaflets was 

 naturally that they represent cases of the development of a typi- 

 cal leaflet in the place of the terminal scale. In common morpho- 

 logical parlance such leaves as these might be spoken of as ata- 

 vistic, the terminal leaflet being considered a reversion tq an 



