LOWER CARBONIFEROUS FOSSILS. 487 



while the upper Madison, 1 represented by 33 species, and the lower Madi- 

 son by 47 species, have 21 species in common. In other words, 64 per 

 cent of the forms found in the lowest beds occur also in the middle beds, 

 while 45 per cent survive into the upper beds; 60 per cent of those in the 

 middle beds are found also in the upper, and 64 per cent of those in the 

 upper are introduced in the lower Madison. These percentages include 

 species which were found in only a single locality, and which, though per- 

 haps alien to the beds in which they are not recorded, at the same time 

 can not be said to be characteristic of that which alone provided them. 

 Although the table (see p. 484) shows a number of species in the upper 

 jjortion of the Madison which have not yet been found in the beds below, 

 it would scarcely be true to say that they materially changed its character. 

 The fauna of the Madison limestone is closely related to that which 

 was described in reports by White, 2 by Hall and Whitfield, 3 and probably 

 also by Meek. 4 The close relationship of this fauna with that of the 

 Kinderhook formation of the Mississippi Basin was recognized by the 

 authors mentioned. Writing in regard to this correlation in 1877, White 

 says (loc. cit.): "The collections of the expedition contain fossils from 

 only three localities that I have definitely referred to the sub-Carboniferous 

 period. These localities are Mountain Spring, Old Mormon road, Nevada ; 

 Ewells Spring, Arizona (upper horizon), and a place below Ophir City, 

 Utah. The collection made at the first-named locality is the most charac- 

 teristic and important one of all, and is referred to the horizon of the 

 Kinderhook formation, to which horizon it is not improbable the others also 

 belong." And again: "The case is far different, however, with the collec- 

 tion from the Mountain Spring locality, which I refer without hesitation to 



1 In making out these percentages certain species, as far as my data go, are found to have skipped 

 one or more beds in vertical distribution. In such case the species have been counted as occurring 

 uninterruptedly between the points of highest and lowest occurrence. 



2 Wheeler's Rept. U. S. Geog. Surv. W. 100th Merid., Vol. IV, 1877, pp. 12-17, p. 79 et seq. 



'King's Rept. Geol. Expl. 40th Par., Vol. IV, 1877, p. 251 et seq. 



1 Hayden, 1873, Prelim. Rept. U. S. Geol. Surv. Wyoming, etc., Sixth Ann. Rept. (for 1872), pp. 

 432-433. Meek here cites a fauna from Mystic Lake; canyon, east sideof Madison River; BridgcrPeak, 

 near Fort Ellis ; Blaektail Deer Creek, north side of Gros Ventres Butte : Flathead Pass, north side 

 Henrys Lake; canyon west of Gallatin River (all in Montana), and Camp 19, Idaho, of which he 

 says (p. 433) : " In looking over the collections from these localities I have been quite impressed with 

 the similarity of their general facies ( without being quite sure that any of the species are identical) to 

 the fauna of the Waverly group of Ohio, now known to belong to the Carboniferous." Most of these 

 localities are in the vicinity of Yellowstone National Park, and probably are in the Madison lime- 

 stone. 



