THE SUDDEN ORIGIN OF NEW TYPES 411 



the elongated digits possess claws ; in the adult Loggerhead 

 {Thalassochelys carettd) only the claw of the first digit remains. 



It is noteworthy that the paddle-finned fishes with Amphi- 

 bian affinities, both the Crossopterygii and the Dipnoi, reached 

 a very great development in the Middle and Upper Devonian, 

 probably as a result of the severe desert-conditions of the Old 

 Red Sandstone, when the fishes of the evaporating seas and 

 lagoons were forced either to become adapted to an amphibian 

 existence or to perish miserably from suffocation and starvation. 

 The exigency of crawling on the mud by means of paddle-fins 

 would soon result in the evolution of an elbow-joint, just as it 

 has done at the present day in the case of the Hopping Gobies 

 (Periophthalmus and Boleophthalmus) of the tropical mangrove 

 swamps or the somewhat similar blenny Alticus, which is also 

 adapted for progression on land. 



One of the clearest instances of the principle set forth in this 

 paper is exemplified by the sudden rise of the Hydrozoan class 

 of Graptolites.^ The oldest form is the branched, tree-like 

 Dictyonema, occurring in fanlike or funnel-shaped colonies. It 

 appeared first of all in the Cambrian of the Appalachian Moun- 

 tains, and — as in many other characteristically Silurian groups — 

 a general eastward migration gradually , took place, so that 

 Dictyonema is found at a somewhat later date in Northern 

 Europe ; in the Baltic Region it is found immediately overlying 

 the Obolus-conglomerate of the uppermost Cambrian. 



Other early forms of the Dendrograptidi {Dendroidcd) or 

 dendroid Graptolites are Dendrograptus {e.g. the dichotomous 

 D. Hallianiis of the Potsdam Sandstone), Callograptus and 

 Ptilograptus ; in all of these branched fixed forms, with 

 uniserial hydrothecae, no virgula occurs. Dictyonema, how- 

 ever, which lasted from the Upper Cambrian to the Middle 

 Devonian, showed remarkable and unusual persistence for this 

 class of organisms, perhaps owing to the fact that it may not 

 have grown on the sea-bottom like its nearest relatives but was 



' It is necessary, however, not to lose sight of the circumstance that the rela- 

 tion of the Graptolites to the Calyptoblastoid Hydrozoa is no longer universally 

 accepted. The appearance of Diplograptus with its central plate (pneumatophore) 

 partly covering a series of vesicles (gonothecae) from the base of which the 

 hydrorhabds radiate outwards has led to the suggestion that the Graptolite colony 

 was a Medusa-like organism, and that the hydrorhabds were its tentacles. 



