THE DEVONIAN PERIOD. 481 



from the land. It may fairly be assumed that when land plants, 

 insects, and terrestrial mollusks are present in relative abundance, 

 and true marine forms are absent, the associated fishes are land- 

 water forms. When land plants, insects, terrestrial mollusks, and 

 fishes are mingled with a limited selection of species that are usually 

 marine, but which may have frequented somewhat freshened waters, 

 the better inference is that the mixture took place in brackish waters 

 largely surrounded by land from which the true land forms were floated 

 in. When land plants, insects, etc., are commingled with a full group 

 of true marine forms, the better inference is that the former were 

 carried out to sea by floods, and that the deposit was really marine. 

 In such an association no safe inference can be drawn relative to 

 the habitat of the fishes, unless there is independent evidence, for 

 they may have been true dwellers in the sea, or may have been floated 

 out from the land streams. If these three cases occurred alone, the 

 interpretation would be fairly easy; but it so happens that in some of 

 the great Devonian series, certain parts of the formation embrace 

 cases of the first or second class, while in other parts cases of the third 

 class are presented. The better inference here is that the formation 

 took place in part on land or in land waters, and in part in marine 

 waters. The deposits forming to-day in the great valley of California 

 embrace fluvial, lacustrine, and marine deposits, together with land 

 aggradation. At a geologically distant day, only small portions of 

 this aggregate would probably be found fossiliferous, and the evi- 

 dences of these portions might be quite opposed to one another, accord- 

 ing to the particular parts selected; but the true interpretation is 

 obvious. This appears to be very much the state of things that is 

 recorded in the " Old Red Sandstone " type of the Devonian deposits. 

 In America, the Catskill deposits, or some large part of them, appear 

 to belong to this type. It is clear that they were contemporaneous 

 with the Chemung in part, and the opinion of Williams and others 

 that in their total range they may even have been contemporaneous 

 with the whole of the marine Devonian series is inherently reason- 

 able, for there can be little doubt that there were lacustrine, fluvial, 

 and land-aggraclational deposits at all times throughout the period. 

 The only question is whether these have been preserved and are now 

 known. In the Canadian provinces, deposits of this kind seem to 

 have a quite ample development. But as all these deposits were, 



