408 
EATON fel ed 
[ Sept. 21, 1871 

as the mandible and the hyoid and scapulary arches. 
From a study of the skull, it kecomes apparent at once 
why in fossil teeth of Cervatodus nothing or very little of 
the bone attached to them has been preserved. Those 
teeth rest on cartilage as well as on bone, the latter being 
a very thin and porous layer which could not be preserved, 
unless the progress of stratification had been going on 
with as little disturbance as in the Solenhofen Schiefers ; 
but the matrix in which fossil Ceratodont teeth are found 
shows that it was formed under very different conditions, 
and it is certainly not of a nature to permit the supposition 
that thin porous lamellze of bone would have been pre- 
served entire. 
The structure of the skeleton reminds us much of that 
of the sturgeons, Chimeera, and especially of Lepzdosiren,; 
and of all the modifications by which it differs from these 
types, perhaps none is of greater interest than that ob- 
served in the paddies. The central part of the paddle, 
which we have found externally to be covered with scales, 
is supported by a jointed axis of cartilage extending 
from the root to the extremity of the paddle ; each joint 
bears a pair of three- or two- or one-jointed branches. 
This is the case in the hind as well as fore paddles, and 
we are justified in supposing that those extinct Ganoids of 
which impressions of paddles with scaly centres have been 
preserved, were provided with a similar internal skeleton. 
Professor Huxley, some years ago, drew attention to 
the analogy existing between the filamentary limbs of 
Lepitios?rén and the lobate fins of certain extinct Ganoids, 
and the correctness of this view is fully borne out by the 
discovery of Cevatodus, inasmuch as the Lepidosiren limb 
proves to be typically the same as that of Ceratodis, but 
reduced to the jointed central axis. 
The gills are perfectly developed, four on each side. 
They are broad lamellated membranes, free from each 
other, but attached to the outer walls of the gill-cavity. 
One can hardly doubt that, in water of normal composi- 
tion, they are sufficient for the purpose of breathing. A 
lung, however, is superadded to them, a true lung, which 
receives blood from a branch of the aorta, and returns it 
directly to the heatt by a separate vein. Whilst the 
Barramunda is in water sufficiently pure to yield the ne- 
cessary Supply of oxygen, the function of breathing rests 
with the gills alone, and the lung receives arterial blood, 
returning venous blood, like all the other organs of the 
body ; under this condition it does not differ from the 
air-bladder of other fishes. But when the fish is com- 
pelled to sojourn in thick muddy water, charged with 
noxious gases, which must be the case very frequently 
during the droughts which annually exhaust the creeks of 
tropical Australia, it commences to breathe air in the way 
indicated above , under this condition the pulmonary vein 
carries purely arterial blood to the heart, where it is 
mixed with venous blood and distributed to the various 
organs of the body. If the medium in which the fish 
happens to be is perfectly unfit for breathing, the gills 
cease to have any function ; if only in a less degree, the 
gills may still continue to assist in respiration. In short, 
the organisation of the Barramunda is such as to justify 
us in the assertion that it can breathe by either gills or 
lung alone, or by both simultaneously. 
With regard to the structure of the lung, it shows a 
nearer approach to the air-bladder of other living Ganoid 
fishes than that of Lefidoszren ; it is not paired, but con- 
sists of a single long sac extending nearly to the end of 
the abdominal cavity. Yet the interior of the sac shows 
a symmetrical arrangement of the right and left side, being | 
subdivided into numerous cellular compartments, by which | 
the respiratory surface is much increased in extent. 
The next organ of importance for determining the 
systematic affinities of the Barramunda is the heart. 
Considering the great resemblance this fish has shown in 
other respects to LePzdosiven, I fully expected to find this 
organ agree also with the Dipnoous type ; but this is not 


the case. Instead of the two longitudinal valves of the 
Dipnoous heart, the du/bus arteriosus is provided with two 
or three transverse series, of which one only is fully deve- 
loped ; or, in other words, Ceratodus proved to be a Ganoid 
fish. But, as Ceratodus and Lepéidosiren are in all other 
points too closely allied to be separated in two distinct 
sub-classes or even sub-orders, we must arrive at the con- 
clusion to drop the Défuoz as a sub-class, and to refer 
Lepidosiven also to the Ganoids, which will then be charac- 
terised, not by transverse series of valves, but by the 
presence of a muscular, contractile dvdbus arteriosus with 
valves, transverse or longitudinal, in its interior—a struc- 
ture which they have in common with the sharks and rays 
(Plagiostomata). ] 
The intestinal tract is a large straight sac wich an 
internal spiral valve, as in the Ganoids and Plagiostomes. 
The kidneys are paired, the ureters enter a very small 
urine bladder or cloaca at the back of, and partly confluent 
with, the rectum. 
The organs of propagation show some noteworthy 
peculiarities. They are paired, inlong bands. The male 
organs have no visible outlet, although a seminiferous duct 
has been found traversing the substance of the testicle 
through nearly its whole length ; no outward opening could 
be discovered, and it is not known how the semen 
is discharged. The ova are small, very numerous, and 
attached to transverse laminz of the ovaries; when 
mature, they fall into the abdominal cavity, as in the 
salmon tribe, and would appear to be expelled through 
two wide slits behind the vent. Yet each ovary is accom- 
panied by a long oviduct, as in the sturgeon or Lefido- 
siven, though it probably has no function, and is only 
indicative of an approximation of this remarkable fish to 
higher types. Such are some of the principle features of 
the organisation of the Barramunda ; and it remains now 
to add some remarks on its affinities and its place in the 
system, A. GUNTHER 
(To be continued.) 

ON EXOGENOUS STRUCTURES AMONGST 
THE STEMS OF THE COAL-MEASURES 
N a memoir recently read before the Royal Society, I 
propounded a new classification of the vascular cryp- 
togams, and at the late meeting of the British Association 
at Edinburgh I brought the same subject forward, when 
my views were opposed by Mr. Carruthers, Dr. M‘Nab, 
and Prof. Dyer, as reported in the columns of NATURE 
for Aug. 31. I was well aware that when I disturbed 
existing and time-honoured systems of classification I 
should meet with such opposition ; but, being thoroughly 
convinced that my views are sound, and that they will 
ultimately be adopted, it only remains for me to face the 
conflict, and persevere with my demonstrations of what I 
believe to be true. My present object is to do what was 
impossible in the hurried and unsatisfactory discussions 
that frequently arose at the meetings of the British Asso- 
ciation to accomplish, viz.: to take care that there shall 
be no misunderstanding as to the real points at issue. 
My opponents seek to interpret the gigantic arborescent 
stems of the coal-measures by the light of the dwarfed 
and degraded examples of vascular cryptogams which 
constitute their living representatives. 1, on the other 
hand, claim to interpret the latter by the former, some of 
which, the Lycopods, for example, instead of being feeble 
things trailing in the grass, had stems three feet in diame- 
ter, and rising a hundred feet into the air. Instead of 
merely constituting a verdant carpet for forests of noble 
exogens and endogens, they were the forest; here, con- 
sequently, we might expect that whatever characteristic 
features they possessed would be developed and displayed 
in their utmost perfection. 
Mr. Carruthers’ fundamental argument is, that I, in my 
