234 



THE AMERICAN NATURALIST [Vol. XLI 



Pennsylvania, West Virginia, North Carolina, and Tennessee. 

 If H. psi/codes is substituted for H. fimbriata in the comparison, 

 the geographic result has the same influence on the discussion. 



H. orbiculaia Torr. and H. macrophylla Goldie are so close 

 that the plants of the two kinds have long been accepted by col- 

 lectors and described by authors as of one species. The differ- 

 ences are at first sight slight, but are apparently constant and 

 sufficient for distinction. The former species has a much greater 

 range, which completely covers that of the latter. II. orbiculaia 

 extends from Labrador and Newfoundland westward through 

 Michigan, and Minnesota, to British Columbia and Washington; 

 and southward through New England, New York, and Pennsyl- 

 vania to South Carolina and Tennessee. It is found in every 

 district where H. macrophylla is found; viz., Newfoundland, 

 New Brunswick, Ontario, Michigan, New England, and New 

 York. The status of these two species is discussed by Ames in 

 Rhodora for January, 1906, with illustrations of the flowers. 



The diflBcult genus Spiranthes has lately been thoroughly 

 studied by Ames, who has given the results in Orchidaceje, Fasc. 

 I, pp. 113-154. The abundance of material examined may be 

 seen from the citations of specimens in the detailed statement of 

 the distribution of each species. 



*S. cernua Rich, has for nearest allies, first, the variety (which 

 some authors regard as a species), S. cernua var. ochroleiica Ames, 

 and secondly the species S. odorata Lindl. S. cernua may be paired 

 with either of them. 8. cernua and S. cernua orchroleuca in the 

 dried state can be separated with certainty by no macroscopic 

 character. They may be distinguished by the seeds, however, 

 S. cernua being polyembryonic.^ Unfruited specimens being 

 indistinguishable in the dry state, the exact distribution of each 

 form may not be very precisely defined, but Rydberg in Britton's 

 Manual gives the range of var. ochroleuca as from New Hampshire 

 and Massachusetts to Pennsylvania and North Carolina. I 

 myself have identified, as being unmistakably typical S. cernua^ 

 specimens from Massachusetts, Ontario, Iowa, and Georgia. 



^ See my notes on the embryology of the two forms in Rhodora 2, p. 227 

 (1900) and 3, p. 61 (1901). In S. cernua embryo formation takes place without 

 poUination. 



