THE SUDDEN ORIGIN OF NEW TYPES 421 



Several zoologists^ have indeed within recent years treated 

 the Trilobites as a generalised central group, from which not 

 only Insects and Myriapods but even Pycnogonids (not yet 

 known in a fossil state) have arisen in divergent directions. 



On the assumption of the descent of Insects from a Trilo- 

 bitan offshoot which had taken to a fresh-water habit, it is 

 at any rate noteworthy that according to the theory of this 

 paper the class of Trilobites can be held to comprise all 

 the necessary characters which have already been seen to 

 be preliminary to the origin of a new group : in particular 

 the body-segments are extremely numerous and vary greatly 

 in number. This circumstance is of course sufficient to initiate 

 a great amount of variability. It is also a significant fact 

 that the last marine survivors of the class occurring in the 

 Carboniferous and Permian belong to the order Proetidcv, which 

 comprises trilobites with the largest number of thoracic 

 segments, viz. up to 22. It is true that so large a number 

 is not reached in the latest (Permian) forms, viz. Phillipsia 

 (9) or in Proetus (8-10), which ranges from the Ordovician to 

 the Permian of the United States - — an unparalleled length 

 of range for a Trilobite — but the Proctidcc were evidently a 

 somewhat generalised type of Trilobite, and it is only from 

 a generalised stock that a new class can be evolved.^ 



It does not seem to be outside the range of probability 

 that the severe desert-conditions of the Old Red Sandstone 

 period, with its evaporating lagoons and shrinking rivers, may 

 have been the means of causing in the (hypothetical) fresh- 

 water branch of the Trilobites the evolution of greatly extended 

 pleurae, first of all for the purpose of offering a greater respira- 

 tory area in waters steadily becoming poorer in oxygen. For 

 this theory it is necessary to assume that the Trilobitan 

 ancestors of the insects and myriapods had abandoned the 



' A. Handlirsch, Die fossilen Tnsekten^ etc. Leipzig, 1906-8. 



^ G. G. Shumard, Report on the Geology of Western Texas. Austin, 1886. 



' The Cambrian Paradoxidce, which probably stand close to the ancestral 

 stock, is the only other order of Trilobites in which the thoracic segments approach 

 so large a figure, viz, 16-20. In the Cheiruridce, which are nearly allied to the 

 Pro'etidc?, the thoracic segments vary from 9-18, and the pleurae become partly 

 separated from each other in Cheirurus and are totally separate in the phantom- 

 like Deiphon ; if these separated pleuree became flattened and membranous they 

 would almost foreshadow tracheal gills, which would (as discussed later) appear to 

 be the forerunners of wings. 



