DESCRIPTION OF THE FORMATIONS 381 



origin and the false-bedding a mark of eolian agency; yet it is clear that 

 false-bedding of a certain type is characteristic of river deposits, whereas 

 that due to dune structure is strikingly different. It is, furthermore, the 

 millet-seed texture of sandstones, the rounding of the very fine sand, 

 which is a distinctively eolian feature, rather than the mere occurrence 

 of rounded grains in a coarse sandstone. There are undoubtedly both 

 fluviatile and marine sandstones which show remarkable cleanness and 

 rounding of coarse sand grains. 



That the Upper Old Red is dominantly fluviatile, not eolian and tor- 

 rential in its nature, is indicated further by the following description by 

 A. Geikie of the Dura Den section in eastern Fife, from which the greater 

 number of the fossil fish from this formation have been obtained : 



We may conclude that in this district, as further to the west, the prevailing 

 rock is sandstone, which in the lower half of the formation is of dull purplish 

 to brick-red colors, while in the upper half, as has been already mentioned, 

 yellow and white tints prevail, though by no means to the exclusion of the red. 

 With the sandstones are associated seams of red marl or clay, in which layers 

 of sandy limestone or cornstone may occasionally be noticed. Here and there 

 thin bands of fine conglomerate may be seen, the pebbles consisting sometimes 

 of well-rolled quartz, sometimes of andesites, from the older part of the Old 

 Red Sandstone. 



In the fish-bearing strata of the Carboniferous formations of this district 

 detached scales and teeth are much more frequent than entire skeletons, and 

 are almost always associated with abundant coprolites and remains of ferns 

 and other vegetation. But at Dura Den the conspicuous feature is the crowd- 

 ing together of entire specimens on the chief ichthyolite platform. The fishes 

 appear to have been suddenly killed in shoals and to have been entombed 

 among the fine sandy silt on the spot, before large numbers of them had time 

 to decay and fall to pieces and before any agitation of the water of the lake 

 could separate and scatter their scales and bones. Various causes have been 

 suggested for such rapid and extensive destruction of life, such as earthquake 

 shocks, the escape of mephitic gases, and the isolation of the animals in pools 

 where the water ceased to be adequately oxygenated. No definite clue, how- 

 ever, has been found either in the strata themselves or in the position or con- 

 dition of the fish-remains to indicate what may have been the agent that in 

 this instance wrought such havoc in the fauna. 



Elsewhere in Fife and in Kinross-shire, though traces of fish-remains are 

 frequent enough in the Upper Old Red Sandstone, they consist for the most 

 part of single scales and fragments of bone, especially belonging to the genus 

 Holopty chius. As they are not restricted to any definite horizon, but seem to 

 range throughout the whole formation, it would appear that the same general 

 conditions of sedimentation prevailed from the beginning to the end. The 

 Dura Den sandstone does not. therefore, so much mark a definite palseonto- 

 logical subdivision as an exceptional area where the organisms were rapidly 

 killed and buried in great numbers. Similar yellow sandstones occur inter- 

 leaved with red bands along the outcrop of the higher portion of the forma- 

 tion westwards all through the Howe of Fife into Kinross-shire. No such 

 fossil riches as those of Dura Den have been met with, however, in these more 



