228 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxi. 



that figure the diftereuces are obvious. Except for its small size it 

 might be compared to C. marylandica from the Iron Ore Clays of Mary- 

 land, and of all the specimens of that species it most resembles the 

 fragment which Professor Foutaine designated as No. 2,' and which I 

 have described as Johns Hopkins Cycads, No. 3.^ That specimen, how- 

 ever, has a large secondary axis, which with better material might 

 take it out of that species. 



Of all the forms from the Black Hills it most resembles iu the char- 

 acter of the scars, etc., some of the smaller branches of C. marshiana, 

 in which these are considerably reduced in size. I have, therefore, had 

 a faint suspicion, which I would not leave the subject without express- 

 ing, that it might be one of these secondary axes or knots, as it were, 

 wrenched from the larger trunk and found in an isolated position. I 

 have, with this thought in my mind, examined a great many such 

 cases, but I can find none in which the fracture at the point of separa- 

 tion at all resembled the base of this specimen, they all showing the 

 break to have been due to some extraneous cause, whereas the base of 

 this specimen is jierfectly natural, not torn nor cracked, and shows the 

 medulla at the center. Nevertheless, there is something a little 

 anomalous in the way the armor surrounds the axis. 



The above arrangement of the species of Cycadeoidea from the Black 

 Hills is not wholly without method. It is true that there is no lineal 

 arrangement that can be regarded as satisfactory, and yet there are 

 decided aftinities among the species. These affinities, however, are 

 shown in particular characters, and the same sj)ecies may have some 

 characters almost in common with two or more other species that are 

 otherwise very difiereut. This is specially the case with branching- 

 species in which other characters resemble those of unbranched si)ecies. 

 For example, 6\ turrita, except in its branching habit, is closely allied 

 to C. mchridei, which never branches; G. marshiana, but for its branch- 

 ing, would be nearly related to C. colossalis ; and C. wellsii may be 

 almost regarded as an unbranched form of G. minnelahtensis. 



In view of these and many other more subtle peculiarities, I have 

 sought, since the arrangement must be lineal, to compromise in such a 

 manner as to bring those species most akin as near together as possi- 

 ble, but it is clear that any arrangement would widely separate species 

 that are similar in one respect or another. 



G. dat'ofcnsi,s and G. colossalis are obviously very closely allied species. 

 G. wellsii can scarcely be said to form a transition from G. colosaalin to 

 G. minneJcahtoisisj but it resembles the former at least in the one fact 

 of being simple. G. pulcherrima is somewhat close to G. minnekalitensis. 

 C civatricula can not be said to form a transition from G, pulcherrima 

 to G. turrita, but it has considerable alfinity to the former. G. turrita 



I Potomac or Younger Mesozoic Flora, Monogr. U. S. Geol. Surv., 1890, XV, p. 192. 

 ^ Proc. Biol. Soo. Washiugtou, March 13, 1897, XI, p. 11. 



