02 The Ann rim i, Geologist. Jan. 1891 
show the complicated enamel folds thai characterize the labyrintho- 
donts. They may be anchylosed to the outer side or to the summit of 
the jaw. or they may be set in distinct sockets, and in many cases they 
are continuously renewed during life. The form of reptilian teeth is 
simple, being for the most part a flattened cone. 
The change which recent research lias brought about in our knowledge 
of this class may be realized from the author's statement on page 1057. 
■'As regards the classification of reptiles scarcely any two writers 
agree. There is little difficulty with existing forms, but when we go 
back to the early part of the Mesozoic era we find that nearly all the 
orders into which the class has been divided show such signs of passing 
into one another that it is quite impossible to exhibit their relationship 
by any system of linear classification." 
In view of this difficulty the author provisionally adopts, with some 
slight change, the classification proposed by Dr. Baur, now of Worcester, 
Mass., in which they are divided into ten orders. Of the first of these 
the anomodonts (Theromora or more correctly Theromorpha of Cope). 
He says the evidence shows almost conclusively that its members are 
descended from the labyrinthodont amphibians and more especially 
from the archegosaurians. 
The difficulty of classification and the increase of our knowledge of 
the phylogeny of the Reptilia is in great part due to the rapid progress 
of discovery in this field. Recent researches, especially in the Jurassic 
strata of the western states have brought to light an immense number 
of forms for which no place could be found in any previous system. 
Many other species are known only from a single bone in some geolog- 
ical museums, and of these several will ultimately perhaps be proved to 
belong to the same animal. 
The shapes assumed by some of these fossil reptiles are most extrava- 
gant. In Dimetrodon the neural spine was according to Cope twenty 
times as long as the centrum and formed an elevated fin, while in 
Naosaurus which is figured by Mr. Lydekker, a similar neural spine 
carries six projecting cross processes on each side. Dicynodon laccrti- 
ceps has the "canine" teeth of a tiger in a skull twenty inches long, 
while its near relative Oudenodon (Udenodon) was toothless. The fa- 
miliar Plcsiosaurus had the long neck of a swan combined with the head 
of a lizard and the paddles of a turtle. Testudo (Colossochelys) atlas 
measured six feet in length and was probably the largest tortoise that 
ever existed. Ichthyosaurus was thirty feet long with the head of a 
crocodile and eye-sockets six inches in diameter. It was carnivorous 
and its coprolites prove that its food consisted in great part of ganoid 
fishes, while the frequent presence of uninjured young in the body- 
cavity tends to prove that it was viviparous in some species at least. 
This genus, the type and almost the only component of its family, has 
not been found in America. Hypcrodapcdon had no teeth but its jaws 
were encased in hard horny beaks like those of a falcon, the upper 
being stronglv hooked. 
