428 



SCIENCE 



[N. S. Vol. XXVII. No. 689 



makes species out of variations or varieties 

 (" forms " or " kinds "). 



Consequently, when Leavitt demonstrates 

 tliat there are many " pairs of kinds " of 

 plants, which possess the same range, he does 

 not prove that the sentence referred to is in- 

 correct, but he may have shown only that the 

 " kinds " he discusses are not species. 



This is clearly seen in the instance of 

 Cypripedium pubescens and parviflorum (p. 

 235). Even granted that the ranges of these 

 two forms practically coincide (p. 236), we 

 may conclude as well that they are not species. 

 This indeed is supported by other observations. 

 O. W. Knight has shown' that these two 

 " forms " are merely variations, which may be 

 observed successively in one and the same 

 plant, according to the character of the en- 

 vironment: 0. parviflorum has been trans- 

 formed into C. pubescens by the change of the 

 environment, and vice versa. 



In other cases, quoted by Leavitt, we may 

 have to deal with species : but the facts given 

 are too scanty to form an opinion. I have 

 repeatedly emphasized that only an exact and 

 complete knowledge of the distribution is apt 

 to give us the means of judging as to segrega- 

 tion and speciation of the forms in question. 

 Leavitt's instances are very defective in this 

 point, and I shall take up one of them, in order 

 to make clear what I mean. 



The range of Spiranthes cernua (= Ihidium 

 ■cernuum) is given (p. 234 f.) as : Massa- 

 chusetts, Ontario, Iowa, Georgia;" and that 

 of Spiranthes odorata as : Georgia, Florida, 

 Alabama, Louisiana, Texas." Thus, as 

 Leavitt says, the range of the latter " coin- 

 cides widely with that of S. cernua." How- 

 ever, we do not know whether the two 

 " forms " are ever found associated, i. e., 

 under the same environment, in those parts 

 of their range which overlap in the southern 



' " Some Notes on our Yellow Cypripediums," 

 Rhodora, 8, 1906, pp. 93, 94. 



' According to Britten and Brown, " Illustrated 

 Flora of the United States," 1, 1896, p. 471: Nova 

 Scotia to Ontario and Minnesota, south to Florida 

 and Louisiana. 



• Ibid. : North Carolina to Kentucky, Florida 

 and Louisiana. 



states. Besides, it is evident that the one is 

 more northern, the other more southern in its 

 distribution, and possibly their centers of 

 origin and radiation were segregated accord- 

 ingly. And finally, the morphological facts 

 are difierent from what they appear in 

 Leavitt's statement. 



According to this (and to the quotations 

 from Britton and Brown), we should expect 

 Ihidium cernuum in western Pennsylvania. 

 But this species seems to be absent here, and 

 it is represented by another one, described 

 recently by O. E. Jennings :" • Ihidium in- 

 curvum. This was first discovered in Erie 

 County, Pa. But additional localities in 

 Beaver, Allegheny, Westmoreland and Arm- 

 strong Counties, Pa., are represented in the 

 herbarium of the Carnegie Miiseum (the speci- 

 mens were partly labeled cernuum, partly 

 odoratum), and the new species has been 

 found in large niunbers very recently (Sep- 

 tember and October, 1907) in Allegheny 

 county by Jennings, and in Beaver county by 

 the present writer. Ihidium cernuum is ab- 

 sent in this part of Pennsylvania : at any rate, 

 all plants called by this name seem to be 

 incurvum. Thus, with reference to Ihidium 

 cernuum and incurvum,, the rule holds good, 

 that two closely allied species are not found 

 associated under the same conditions, and this 

 rule also fits beautifully the case of Ihidium 

 incurvum and odoratum, which are even more 

 closely allied. It remains to be seen what the 

 distribution of 7. incurvum will turn out to 

 be outside of western Pennsylvania. Possibly, 

 many plants called cernuum or odoratum may 

 be the new species. 



Of course, the question is not yet settled. 

 I only took up this instance in order to show 

 how incomplete our ■ knowledge is. Large 

 tracts of the country are poorly or not at all 

 known, and in addition, the knowledge of the 

 morphological details is often defective. It 

 would not be astonishing, if finally it should 

 be found that Ihidium odoratwn runs through 

 I. incurvum into I. cernuum,, but for the pres- 

 ent all these forms are morphologically well 

 separated, a separation which corresponds, at 



^"Annals Carnegie Museum, 3, 1906, p. 483. 



