scHrcHERT] DEVELOPMENT OP THE TELOTREMATA. 87 



at different times aud in separate phyla straight cardinal areas are 

 more or less well developed. In America, the oldest jnembers of this 

 order {Protorhj/nchaf minor and P.f amhigua, members of the family 

 Rhynchonellid.e) occur in the Lower Cambi-ian. In these species, and 

 in the great majority of this family, there is no cardinal area; but 

 occasionally this character is present, the earliest conspicuous example 

 being the Ordoviciau genus Orthorhynchula. Among the Paleozoic 

 Terebratulacea cardinal areas are seldom developed. A conspicuous 

 exception, however, occurs in Tropidoleptus. But in the Mesozoic 

 and Ceneozoic, in the family Terebratellidte, cardinal areas are very 

 often present, and in living forms are accompanied by a short pedicle. 

 It is, moreover, in the Spiriferacea, the youngest superfamily of the 

 Telotremata to originate, that the greatest development of cardinal 

 areas takes place. The oldest genera of the Spiriferacea are all ros- 

 trate, as in the Ordoviciau Zygospira, Catazyga, and Cyclospira. In 

 the Silurian the Spiriferidie tend to develoi) rapidly long, straight, and 

 wide cardinal areas, attaining greatest development in the Devonian 

 and early Carboniferous. This excessive development of cardinal 

 areas is no doubt due to the shortening and decline of the pedicle, 

 since in the Triassic system forms occur in which cementation is com- 

 plete (Zugmeyeria and Thecocyrtella). Cardinal areas are also devel- 

 oped in other families of the Spiriferacea, but in no case can such be 

 traced to Ordoviciau long-hinged ancestors. 



In this order, more than in the Protremata, internal specialization of 

 the brachia has progressed from a simple to a highly complex condi- 

 tion, hi the Protremata, in its latest developed superfamily, Penta- 

 meracea, crura are a,lso ])resent, of the same phase of development 

 attained by the Ilhynchonellacea, the most primitive superfamily of 

 the Telotremata. In this order, however, there are, with but few 

 exceptions, no internal special structures, as spondylia. The si)ecial- 

 ization in the Telotremata is expressed in the progressive complica- 

 tion of the calcareous brachial supi^orts. In the most primitive spe- 

 cies of the lihynchonellacea no crura are present (Protorhyncha), but 

 in all later forms these appendages are well developed, and finally in 

 the Trias and Jura attain very great length in Ehynchonellina. In 

 the next more complicated superfamily, Terebratulacea, the crura in the 

 primitive members have united anteriorly, thus forming the simple 

 unchanging loop of Centronella aud Rensselferia, which is also known 

 to occur in the very young of some species of the highest sui)erfamily, 

 the Spirileracea. The geological history of the loop has shown that 

 the brachia have been constantly changing, causing more or less com- 

 lilete resorption of the hard ])arts and adaptation to later requirements. 

 The progressive development of the loop is also repeated ontogenetic- 

 ally and more or less fully in living terebratuloids. 



In Zygospira, the oldest known genus of the suborder Spiriferacea, 

 the primitive loop of Centronella is reproduced in the earliest phase in 



