202 



on the stem of two kinds, those of the lower plane alternate, 

 spaced, spreading, apparently rounded at apex; leaves of upper 

 plane, small (one-third the length of the others), appressed, 

 oblong, obtuse; branchlets all fertile, apparently square, sporo- 

 phyls all alike, alternate, close, overlapping, slightly setose- 

 pointed; sporangia large, one in the axil of each bract, circular 

 (now flattened), ornamented with numerous minute, circular 

 elevations. 



Types. — U. S. National Museum, Nos. 35,009, 35,010, 35,011. 



Locality. — Northeast Montana just below international bound- 

 ary (sec. 33, T. 37 N., R. 47 E.). 



Horizon. — Fort Union formation (Eocene), from clay above a 

 large vein of coal and about 40 feet above the base of the for- 

 mation. 



The working out of the probable affinity of Selaginella Col- 

 lieri among the living species is a matter of some difficulty. As 

 is well known, the genus is a vast one, comprising between 500 

 and 1,000 species, and as many of the species are very close 

 together, and the individual variation often very considerable, 

 an expert knowledge of the genus which the writer cannot lay 

 claim to is really required to stabilize identifications. In the 

 first place it may be pointed out that the fossil species appears to 

 have little or no close relationship with any form now living in 

 the United States or Canada, the nearest species being perhaps 

 Selaginella apus (L.) Spring, but this is obviously so remote 

 as to be negligible. 



In general 5. Collieri appears to approach most closely to 

 S. stenophylla A. Br. {S. Lycnuchus Tourn., Fil. Mex., p. 148, 

 non Sprieng), and especially its variety rigidiuscula Tourn., a 

 native of Mexico. In some features it also resembles 5. didy- 

 mostachya (Desv.) Sprieng, and S. confusa Sprieng, both natives 

 of Jamaica. It also suggests in the form of its branching 5. guate- 

 malaensis Baker, from Guateinala. However, its small size, 

 very close, densely fertile spikes and spreading leaves serve to 

 distinguish it at once from any of those mentioned above, and it 

 is probable that its real affinities have not been determined. It 

 is hoped, however, that the excellent figures given may ultimately 

 serve to allocate it. 



