428 

NATURE 
[ Sept. 28, 1871 

bestowed a great deal of care and trouble, trusting only 
to averages of many repeated measures, did not agree 
satisfactorily with the maker’s statement. I do not know 
whether it may have been generally noticed, but the 
remark is a very obvious one, that the limit of numerical 
error increases with the power, so that in the case at any 
rate of ordinary dynameters, minute accuracy in the 
estimates of very deep oculars bears evidence of its own 
futiliry. If it :epresents anything of value, it can only 
be the care and attention with which a mean was 
deduced from repeated trials; but even this would be 
better expressed in 10und numbers, as less likely to 
convey an erroneous impression to the uninitiated. 
Probably some form of apparatus may yet be devised 
which may secure greater minuteness in the measure- 
ment of very high powers, without entailing a dispropor- 
tionate outlay. In the mean time Mr. Berthon’s invention 
may be safely recommended as likely to prove of especial 
advantage to observers in general, 
T. W. WEBB 


THE NEW GANOID FISH (CERATODUS) 
RECENTLY DISCOVERED IN QUEENSLAND 
Il. 
ie appears to me that there is not the least justification for 
separating the living fish genericad/y from that extinct 
form, the teeth of which are known by the name of Cera- 
todus. Immediately after its discovery became known, 
and before we knew more than the outlines of its external 
characters, views to the contrary were expressed, evidently 
based on the assumption that a genus was not likely to 
have survived from the Triassic epoch. This is certainly 
a remarkable fact, but it is not more surprising than the 
other, viz., that fishes from one of the oldest epochs from 
which fish remains are known, are most closely allied to 
Ceratodus. We know that the same sfeczes occur on both 
sides of the Central-American isthmus—that is, that they 
have existed at, and remained unchanged from, the time 
when the Pacific and Atlantic Oceans were connected by 
one or several channels. Therefore, it would appear that 
there is a greater persistence in the ichthyic type than we 
have hitherto been willing to admit. 
Whoever has compared the teeth of Ceratodus runci- 
natus {rom the German Muschel-Kalk, and those of the 
Indian species described by Prof. Oldham, with the teeth 
of the living species, must admit their generic identity ; 
and if the Australian form really grows to the enormous 
size stated by some colonists, I have no doubt that the 
teeth of such large examples cannot be distinguished 
from the fossils mentioned. So close a resemblance in 
highly specialised teeth like those of Cevatodus is generally 
admitted to be of generic significance. Unfortunately 
no other part of Ceratodus is known to have been pre- 
served in the fossil state to serve as a further guide in 
answering this question. The strata in which the teeth 
‘are found must have been much disturbed, as no two 
teeth have ever been met with z7 sz/w together ; but I 
cannot help thinking that soorer or later the vomerine 
teeth will be recognised. From their smaller size they 
would easily have escaped observation ; and their shape 
(which differs so much from that of the molars) would 
scarcely have allowed an observer to connect them with 
the genus to which they in reality belong. 
The next most nearly allied forms are the American 
and African Lefidosirens, a genus at present unknown in 
a fossil state. The skeleton (in some respects even to its 
minute details), structure of the fins, dentition, internal 
nostrils, three-chambered heart, co-existence of a lung 
with gills, intestinal tract, small size of ova: all present 
the strongest possible evidence of the close relationship 
between these fishes. The points in which they differ are 
of such a nature that characters indicative of an amphibian 


affinity in Lepzdosiren are modified in Ceratodus according 
to a distinctly ichthyic type, thus tying, as it were, Lepz- 
dosiren firmly to the class of fishes. The longitudinal 
valves in the bulbus arteriosus of Lefidoszren, reminding 
us of a similar structure in the heart of Batrachians, are 
replaced by truly Ganoid valves in Ceratodus; the im- 
perfect gills of the former genus are as perfect in the 
latter as in any other fish; the lungs of Lefiédosiren, 
paired as in a frog, are confluent into a single air-bladder- 
like sac in the Australian form ; instead of the closed 
ovaries with a developed oviduct and fallopian tubes of 
Lepidosiven, we find the ovaries of the Barramunda open, 
discharging the ova into the abdominal cavity, as in the 
Salmon family and other fishes. These differential 
characters may be considered by some of sufficient im- 
portance to refer the Lefzdoszvens and the Barramunda to 
two distinct groups. 
Some of the oldest fishes, known from the Devonian 
epoch, are designated by the names Crenodus and Dipterus. 
Whether they should be referred to one genus or two is 
a question about which opinions are divided, and into 
which J need not enter here. They are evidently repre- 
sentatives of the same ichthyic type as the Dipnoi of the 
present epoch. The similarity of the large molars to 
those of Ceratodus has been recognised for a long time ; 
but it is only recently that I have been able to ascertain, 
in an example in the Jermyn Street Museum, the presence 
of a pair of small vomerine teeth. Moreover, the same 
example presents as good evidence as we can expect ina 
fossil, that the nostrils are placed within the mouth. These 
characters are combined with the presence of acutely 
lobed paddles, and of a notochordal skeleton ; but there is 
the great difference that the end of the vertebral column 
is heterocercal, instead of being diphyocercal, as in Lepz- 
dosiren and Ceratodus. Therefore Ctenodus will form 
the type of a distinct dipnéous family. 
Thus, then, we arrive at the conclusion that Lep/dosiren, 
far from being an isolated representative of a distinct sub- 
class of fishes, is only one of the representatives of a sub- 
order of Ganoid fishes, characterised by the position of the 
nostrils within the mouth, by paddles supported by a 
jointed axis, by lungs co-existent with gills, by a noto- 
chordal skeleton, and by the absence of branchiostegals. 
The term “ Dipnoi” may be retained for this sub-order, 
which was developed in the earliest epochs from which 
fish-remains are known, while we have scanty evidence of 
its presence in Liassic and Triassic strata, and, in the 
present state of our knowledge, it appears to be lost, until 
we find it again represented by three living forms in the 
present period. Probably some other extinct genera be- 
longed also to this sub-order, but their remains are in too 
fragmentary a condition to admit of an exact definition of 
their affinities, 
During the examination of Ceratodus,I had so frequently 
occasion to refer to structural peculiarities of the P/agzo- 
stomata (Sharks and Rays), that I wasinduced to reconsider 
the relations existing between this sub-class and the 
Ganoid and Teleosteous fishes ; and I came tothe conclu- 
sion that the two former are much more nearly allied to 
each other than the Ganoids are to the Teleostei. The 
Plagiostomes were considered to be a distinct sub-class of 
fishes on account of the highly-developed state of the 
organs of reproduction in the female, besides the presence 
of copulatory organs in the male. Their ova are different 
from those of other fishes, having a very peculiar shape, 
and shell with adhesive appendages, and being of an un- 
usually large size, and few in number. They are impreg- 
nated internally; some of the species are viviparous. 
They have from five to seven external‘gill-openings. Al- 
though in external appearance a Ray and a Shark are 
apparently very different, yet these extremes are connected 
by a number of intermediate forms, and they form alto- 
gether one of the most homogeneous groups in the zoo- 
logical system. 
