1926] Babcock-Leslei) : Chromosomes and Taxoiiomic Eelalionships 335 



Crepi'S senecioides Delile, a native of Egypt, is a species of peculiar 

 interest because its fruit is definitely flattened, altliough not so much 

 so as in the more extreme types of Lactuca, and it lacks the thin lateral 

 margin (fig. 3, /, /'), while on the basis of its involucre, number of 

 florets per head, and habit it does not fit into any of the sections of 

 Lactuca provided by Ilofi^mann in the Pflanzenfamilien. Further- 

 more, it has four pairs of small chromosomes and produces sterile 

 hybrids when crossed with C. parviflora and C. vesicarm. Thus we 

 find fairly close relationship between what simulates Lactuca in achene 

 shape and certain species of Crepis. This evidence is not uniciue, how- 

 ever, as there are other points at which the two genera meet. Nakai, 

 for example, found it necessary to choose between the alternatives of 

 either recognizing Ixeris, Paraixeris, and Crepidiastrum as distinct 

 genera or combining Crepis and Lactuca. For the present, we are 

 inclined to consider C. senecioides as Crepis, but it is highly desirable 

 that critical comparison of the fruits be made between senecioides and 

 similar Crepis species as well as between senecioides and the North 

 African species of Lactuca, and that chromosome counts of the latter 

 be obtained. We have indicated one such comparison in the drawing 

 of C. hiirsifolia (fig. 3, g,(f). 



A group of forms which have usually l)een treated as distinct 

 species, viz., Crepis vesicaria L., C. taraxacifolia Thuill., C. Marschallii 

 F. Schultz, and C. myriocephala Coss. et DR., may be considered as 

 one species for the following reasons: (1) They are closely similar 

 morphologically, and their close relationship has been recognized by 

 several taxonomists. (2) They have nearly identical chromosome 

 groups. (3) They intercross freely and produce highly fertile hybrids. 

 That these should be considered as subspecies of one species rather than 

 as varieties is indicated by the following facts: (1) All except one. 

 taraxacifolia, which is probably the oldest phylogenetically, occupy 

 distinct geographic areas. (2) All are highly variable, and taraxaci- 

 folia is really polymorphous. However, as no changes in nomenclature 

 are proposed in the present paper, we shall continue to use the 

 binomials in what follows. 



A summary of the data recently presented by Bleier (1925) and 

 Karpetchenko (1925) shows that in Trifolium section Chronosemium* 



* Greene (1897) discusses at length the evidence for retaining the genus 

 Chrysaspis instead of treating it as af section (Chronosemium) of Trifolium. 

 He says: ''And since Linnaeus' time there have been a numl)er of open protests, 

 and by most able botanists, against the treating of the Hop Trefoils as con- 

 generic with such plants as Trifolium praiense and its allies. Systematists of 

 no less renown than Lamarck and Desfontaines referred the plants to Melilotus 

 rather than Trifolium." 



