if^i7] Goodspeed : Notes on Trillium 83 



DISCUSSION 



There seems to be little doubt that within the Eastern species 

 grandiflorum there is a widely distributed form or variety which ex- 

 hibits correspondingh' rather marked variations in the various locali- 

 ties from which it has been reported. This form, sometimes referred 

 to as var. variegatum Pk. (cf. Peck, 1888), is characterized by long- 

 petioled leaves, striped perianth segments, and often an elongation 

 of the peducle. Apparently no other American species possesses any 

 similarly well-marked and widely distributed form or variety based 

 upon abnormal structural variations. Color variations in most of 

 the species are striking and well recognized but probabh' represent, 

 with possibly one or two exceptions, extremes of the normal fluctuating 

 variation of a highly variable genus. Although var. giganteum con- 

 tains no characteristic form of extreme or teratologieal structural vari- 

 ation, the extent and intensity of such variation seems remarkable. 

 That practically all organs of the flower may exhibit duplication of 

 parts, suppression or extreme modification and metamorphosis, has 

 been indicated in the description of specimens above. 



The question of the recurrence from the rootstock and, in the case 

 of perennial plants in general, of abnormal shoots, and the more im- 

 portant question as to the origin of such inherently abnormal root- 

 stocks, are of maximum interest and significance. The genus Trillium 

 would seem to furnish rather imusually valuable and readily investi- 

 gated material for such studies. As yet, however, the evidence at 

 hand, if not contradictory, is at least highly fragmentary and in gen- 

 eral rather unsatisfactory so far as this genus is concerned. The 

 observations on Eastern species, especially those of Deane on un- 

 dulatum, indicate that, in the case of well-defined and extensive tera- 

 tologieal variation, the abnormal condition may be expected to recur 

 year after year in the shoots produced by a given, inherently abnormal 

 rootstock. Our garden cultures of var. giganteum seem to suggest 

 much the same condition, although the more important evidence will 

 come from the results of the next few years. There is no doubt, how- 

 ever, that such apparently relatively simple variations in structure as 

 the change from the trimerous to the tetramerous condition do not 

 recur in succeeding years from the same rootstock. Our cultures 

 appear to indicate that such changes may be environmentally deter- 

 mined, a fact further suggested by the ease with which many of the 



