THE SUDDEN ORIGIN OF NEW TYPES 425 



within the tracheal gills. Of all modern insects the Ephemerids 

 seem to stand nearest to the Carboniferous Palasodictyoptera, 

 and are connected with them by a practically continuous 

 palseontological succession of intermediate forms. Hence it 

 is not unreasonable to presume that some of the modern forms 

 would still possess some relics of archaic organisation to 

 throw light upon the process of the transition from gills to 

 wings, which must have taken place in pre-Carboniferous 

 times. Now as a class the Ephemerids contain more forms 

 with tracheal gills than any other class of insects. At this 

 stage, however, it is necessary to point out that some confusion 

 has arisen from not recognising that tracheal gills may have 

 two distinct modes of origin, viz. (i) lamellar tracheal gills, as 

 in Chloeon (fig. 10), Oligoneuria (fig. 11), Tricorythus (fig. 15), 

 etc., which are clearly modified pleurae, and it is only from 

 these structures that wings could be derived ; and (2) fila- 

 mentous tracheal gills, which are modified abdominal limbs, as 

 in Sialis (fig. 12) and even more clearly in the Hemerobiid 

 Sisyra (fig. 13) inhabiting the cavities of the fresh-water sponge. 

 Now if we consider a section through the Trilobite Triarthrus 

 (fig. 14), it can be seen that there is no inherent difficulty 

 in deriving the lamellar gills of Ephemerids from the pleurae 

 of a Trilobitan ancestor, whilst the fringed and jointed tracheal 

 gills of Sialis can be derived with but little change from the 

 setose exopodite of the Trilobite, and the more leg-like ap- 

 pendages of Sisyra from the endopodite. In the nymph of 

 Tricorythus (fig. 15), a pair of filamentar gills are present 

 at the same time with four pairs of lamellar gills, of which 

 the anterior pair is greatly enlarged to act as a gill-cover 

 to the three posterior pairs and is furthermore nearly equal 

 in size to the immature wings. 



It is unfortunate that the strata of the Old Red Sandstone 

 are so ill-adapted for the preservation of delicate chitinous 

 remains, as the Carboniferous insects already show a very 

 considerable differentiation and reduction to an optimum, e.g. 

 to three pairs of legs and to ten abdominal segments. Evidence 

 is not wanting to show that the reduction had been of very 

 recent occurrence, for the eleventh segment of the abdomen and 

 the telson are not yet quite suppressed ; the wings are equal 

 in size and show similar neuration ; a wing-like expansion with 

 veins is present on the pre-thorax of several Protephemerid 



28 



