1 72 



Thus a total of about 75, certainly not more than 100, hybrids 

 in Hemerocallis have been reported to date. About 40 have been 

 given varietal names but only about a dozen are to be had in the 

 trade. The number available in the American trade is still less. 

 When compared with such groups of plants as the dahlias, the 

 roses, and the irises the flowers of the available day lilies taken 

 collectively are lacking in diversity of color and form. Their 

 improvement along these lines has been the chief aim of the 

 breeding work undertaken at The New York Botanical Garden. 



The first step in preparation for this breeding was to assemble 

 all the known species and varieties obtainable. Persistent efforts 

 have been made to secure the hybrids already produced as noted 

 above but only 14 of these have been obtained at this date. Con- 

 tinued efforts are being made to obtain a more complete collection 

 and also to obtain living plants growing wild in the interior of 

 China in the hope that forms and perhaps species hitherto un- 

 known to Europe and America may be had. Without doubt types 

 of day lilies of interest and value to horticulture are yet to be dis- 

 covered in Asia. During the past year plants have been obtained 

 from China and from Japan which are different from any of the 

 older species. These are being propagated and utilized in the 

 breeding work. 



The main task in the breeding work itself is to obtain seed and 

 to grow seedlings particularly from crosses between dift'erent 

 species and varieties but from self-pollination as well. When this 

 is attempted it soon becomes evident why so few seedlings have 

 previously been secured and why with the probable exception of 

 the Strasbourg hybrids the " blood " of H. fulva has never been 

 blended with any of the other species previous to the results re- 

 ported in this paper. 



The (lay lilies as a group possess a decided unfruitfulness or 

 sterility. It is difficult and for some varieties even impossible to 

 obtain seed to self-pollination. The majority of seedlings are 

 fulK- self -incompatible. Crossing between s])ecies and varieties 

 and e\en between seedlings of the same immediate parentage is 

 greatly restricted. For exani[»lr, the widely cultivated strain of 

 the cd'.nmon orange day lily ( //. fulva) has, it appears, never been 

 known to \ield fruit and seeils to any kind of pollination, self or 

 cross, until four capsules ((tut of sevi-ral thousand pollinations) 



