[Vol. 9 

 102 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



should be lacking in America of some of these less well-known 

 forms. In spite of efforts extending over a considerable interval 

 of time, it has not been possible to obtain adequate material of 

 the following described species: /. Welwitschii, aequinoctialis, 

 Heldreichii, nigritiana, dubia, olympica, Gardneriana, elatior, 

 Muelleri, humilior, triquetra, echinospora var. asiatica, tripus, 

 Savatieri, neoguiniense, baetica, natalensis. 



Wherever possible these species have been placed in the keys 

 according to the characters found in the literature. The species 

 descriptions are based either entirely on the original description, 

 directly quoted or translated, or on that plus supplementary 

 material in later accounts. It has been a matter of concern to 

 choose what seemed authentic. Unfortunately, the monograph 

 of Motelay and Vendryes, which should be of great aid here, has 

 so many typographical errors that one lacks confidence in its 

 reliability. 



KEYS 



In the preparation of keys, the attempt has been made to use 

 as evident characters as possible, such as plant size, character of 

 leaf, and similar features. Unfortunately such points are depen- 

 dent to some extent on external conditions which may prove 

 variable. An effort has been made to take such variations into 

 account; obviously, some errors in judgment, especially because 

 of the use of dried material apart from its habitat, will occur, 

 which in some degree render the key inaccurate. Measurements 

 might better, therefore, be interpreted as suggestive of the prob- 

 able range rather than as absolute limits. 



Furthermore, it has been necessary to utilize minute characters, 

 some of which are not readily determined, particularly in dried 

 material. Previous investigators, as Braunj and Engelmann, 

 have stressed this point adequately. The latter further empha- 

 sized the skill that even the amateur can develop with the op- 

 portunity to study fresh material of several forms, so that stan- 

 dards of comparison may be developed. A little more care is re- 

 quired with the use of dried plants, which involves soaking and 

 more careful methods in dissecting and sectioning. 



Again, with these minute characters, such as ligule and degree 

 of velum development, it is very difficult to more than approx- 

 imate an accurate description. In regard to such a feature as 



