170 



There is little need to give any detailed description of this plant 

 except to call attention to the fact that the five corolla segments of 

 each of the flowers examined were separate to the base. This con- 

 dition was obvious even in the bud (figur£ lb). As in the case of 

 the plant noted by \\"eatherby (loc. cit.), the anthers apparently 

 were abortive (figures Id, e). Whether the sterility extended to the 

 egg-apparatus has not been determined, although it is our observa- 

 tion that this plant does not set fruit. Furthermore, attempts to 

 produce fruit through the medium of artificial pollination have been 

 unsuccessful. However, this last is not a final conclusion, the at- 

 tempts so far having been attended by conditions which admittedly 

 were not ideal. In brief, we are not as yet convinced that this clone 

 is incapable of setting fruit. 



In addition to the polypetalous condition, one other abnormality 

 should be noted. In sectioning the hypanthium of a series of the 

 flowers of this clone it was found that the ovary of each had but 

 four carpels (figure If), instead of the five carpels normal for V. 

 torreyanuin and its close relatives (figure Ij). This, however, is by 

 no means unusual in the genus Vaccininni. It is quite common in 

 certain species and is, in fact, a standard character of others. 

 Nevertheless, this does indicate that the disturbance resulting in the 

 polypetalous condition can also influence the number of carpels in 

 the ovary. In this connection, it is of interest to note that various 

 of the nearly polypetalous members of the subgenera Oxycocciis 

 and Oxycoccoides are tetramerous, with the pentamerous condition 

 being the abnormal form. Whether this condition is merely coin- 



after further and much needed cytological studies of these entities have been 

 made throughout their entire distributions. In addition to the foregoing, the 

 "vacillans-complex" contains V. tallapusae (Cov.) Uphof, a derived tetraploid 

 of the southern Appalachians which is best developed in Georgia; V. alto- 

 ino}itanu)n Ashe of the southern Appalachians may also be a derivative of 

 this group. V. vacillans var. crinitiim Fernald, with which V. niissoiiriense 

 and V. viride have been confused, appears to be a series of hybrids and 

 ecologically selected segregates from crosses between the markedly different 

 V. torreyamim and V. atrococciim, both of which are diploid (»= 12) and 

 known to be interfertile. The ''high-bush" diploid V. atrococcum (A.Gr.) 

 Heller apparently does not enter the ranges of V. missouriense and V. viride, 

 being primarily an east-coast species ; westward, it has been confused with 

 V. arkansanitm Ashe, and with the "arkansanoid" members of the tetra- 

 ploid V. corymbosum L. 



