330 NATURAL SCIENCE [November 
disappear from the records of the rocks; and practically nothing is 
known of them until the early Tertiaries, when the various genera 
and families are almost identical with those surviving at the present 
day. The only satisfactory specimen of intermediate date is a solitary 
skeleton (Hylacobatrachus) from the Wealden of Bernissart, Belgium, 
which seems to represent an animal with persistent gills related to 
the existing Proteus and Menobranchus. 
The story of the early mammals is almost similar. In rocks 
dating back to the close of the Palaeozoic and the dawn of the Mesozoic 
period, there are abundant remains of the Anomodont reptiles or 
Theromora, which make an extremely close approach in their skeleton 
to the warm-blooded quadrupeds which we term mammals. They 
are found in South Africa, India, Russia, Switzerland, Scotland, North 
America, and South America. They must thus have been almost 
world-wide in their distribution. It is also clear that many of them 
attained a very large size. In all the regions mentioned, however, 
they completely disappear above the Trias; and the only known 
Mesozoic fossils which can be referred to the Mammalia are some 
fragments of animals no larger than rats from the Jurassic of England 
and the Jurassic and Cretaceous of North America. It seems, indeed, 
as if the mammals were evolved in some region of the southern 
hemisphere which is either now submerged or not yet geologically 
explored; for they suddenly appear in great numbers and variety 
at the base of the Eocene Tertiary both in Europe and North 
America, which must be the result of migration on the re-arrange- 
ment of land and sea. It is very curious that notwithstanding the 
numerous examinations of the Mesozoic and Tertiary strata of Austral- 
asia and South America during the last half-century, not a single clue to 
the solution of the problem has hitherto been obtained. As suggested 
by Mr Lydekker in a recent issue of the Zransactions of the Geo- 
logical Society of South Africa, it is extremely probable that we must 
turn to a geological exploration of the Dark Continent for the next 
important advance in our knowledge of the subject. 
Our ignorance of the early land-mammals is strange, but the 
want of all knowledge of the ancestors of the marine mammals— 
whales, porpoises, and sea-cows—is still stranger. It is well known 
that at present all the great Mesozoic marine reptiles of the orders 
Ichthyosauria, Plesiosauria, and Mosasauria or Pythonomorpha, seem 
to disappear suddenly at the top of the Cretaceous formations; while 
the marine mammals of the orders Cetacea and Sirenia take their 
place as suddenly towards the top of the Eocene strata. This 
happens not only in Europe and North America, but also in 
Australasia, perhaps likewise in South America. Now, we are 
well acquainted with marine deposits, both of littoral and deep 
water origin, of intermediate age in many parts of the world. 
