368 safford: notes on dahlia 



Dahlia imperialis) or attempting to connect the "duplex" and 

 double forms with their primitive single ancestors. It is very- 

 probable that the types upon which several species have been 

 based were hybrid plants. Dahlia pinnata itself, the type of the 

 genus, was probably a hybrid. In the Index Kewensis its name 

 is discarded as a synonym for the subsequently described D. 

 variabilis. In the same way the handsome Dahlia juarezii with 

 large double heads composed of strap-shaped florets having their 

 edges turning backward, in sharp distinction to the involute or 

 quilled florets of the artificial-looking "pompon dahUas" and 

 the broad, flat-rayed heads of the "century" type of modem 

 catalogues, is also to be regarded as a hybrid. DahUas with flow- 

 ers identical in form with the type of Dahlia juarezii, the ancestor 

 from which the "cactus dahlias" of our gardens have sprung, 

 are no longer called "cactus dahlias" by specialists, but "cactus 

 hybrids." One of the ancestors of Dahlia juarezii must have 

 been a single flowered species, with eight revolute ray-florets. 

 Such a plant has recently been discovered in the mountains of 

 Guatemala by Mr. Paul Popenoe, in honor of whom it is pur- 

 posed to name the species described below. In addition to this 

 species Mr. Popenoe brought back with him a handsome tree 

 DahHa, already represented in the U. S. National Herbarium, 

 but hitherto erroneously referred to Dahlia imperialis by some 

 authorities and by others to Dahlia variabilis. This second species 

 represented in the herbarium by two sheets collected in Guatemala 

 by Mr. WiUiam R. Maxon, is described below under the name 

 Dahlia maxonii. 



There are several other undescribed species of the genus 

 Dahlia in the National Herbarium, but there is no space within 

 the limits of this paper to describe them. Indeed the whole 

 genus should be carefully revised by a botanist familiar with 

 closely allied genera of composites and the work should be based 

 upon material collected in the elevated regions of Mexico and 

 Central America where the plants are endemic, not upon garden- 

 grown specimens. Much of the material in herbariums is in- 

 complete, owing to the absence of characteristic lower leaves 

 of the plants represented; and many of the specimens are in 



