ba 
346 
The ‘ shark recently discovered in Japanese waters,’ 
described by Mr. Garman as Chlamydoselachus an- 
guineus (in Science for Feb. 1, vol. iii. pp. 11€, 117; 
Bull. Essex inst., vol. xvi.), as its describer has re- 
marked, ‘‘is a form of more than ordinary interest 
on account of the respects in which it differs from the 
majority of its kindred.’’ It not only appears as a 
new element in selachology, and becomes the repre- 
sentative of a hitherto unknown type, but it throws 
light on the ancestry and some of the extinct forms 
of the class; and, still further, it may serve as a guide, 
for the interpretation of certain of the tales of the 
sea-serpent. 
In respect to its place in the system, I perfectly 
agree with Mr. Garman, that it is the representative 
of a very distinct family (Chlamydoselachidae): Iam 
also of the opinion that it may be regarded as the 
type of a distinct sub-order at least. Mr. Garman, in 
Science, was ‘inclined to consider this the type of a 
new order, to which the name Selachophichthyoidi 
might be given;’’ but in his article in the Essex 
bulletin he is entirely silent on the subject of the 
major relations of the new type. The name, having 
been thus never defined, and being objectionable on 
account of its length and cacophony, might be re- 
placed by a shorter one, like Pternodonta; but on 
this I shall not insist. A more important question 
is, What is the status of the selachian in classifica- 
tion? Mr. Garman thinks that ‘it stands nearer the 
true fishes than do the sharks proper.’ Idonot know 
how he would express this idea in a linear arrange- 
ment, but most would do so by placing it immedi- 
ately between the selachians and fishes. I am also 
disposed to consider Chlamydoselachus to stand 
‘nearer the true fishes than do the sharks proper,’ 
not because it appears to be in the line of descent 
between the two, but because it is nearer the primi- 
tive line from which both types have diverged. 
Judging from Mr. Garman’s remarks in the two 
articles referred to, I presume there would be essen- 
tial concordance between us as to this point. 
As to the relations of Chlamydoselachus to extinct 
types, however, I must dissent from Mr. Garman. 
Fortunately, an article throwing light on the affinity 
of Cladodus has been published recently, — probably 
too recently to be available to Mr. Garman. I refer 
to Dr. R. H. Traquair’s communication ‘on a new 
fossil shark,’ inthe Geological magazine for January, 
1884 (decade 5, vol. i. pp. 8-8, pl. 2). Dr. Traquair 
has therein made known the form of the cladodont 
selachians, and proved beyond doubt that the clado- 
dont dentition and ctenacanthoid spines co-existed in 
the same fish. The ‘newshark’in which these parts 
were coincident has been named Ctenacanthus costel- 
latus. In the words of Dr. Traquair, “accepting the 
fish just described as a new species of Ctenacanthus, 
it yields us the following important facts regarding 
the genus: — 
'**1, The shape of the animal was moderately 
elongated, with blunt snout and heterocerceal tail. 2. 
The skin was covered with shagreen granules, mostly 
of an ornate, ridged, pectinate character. 38. There 
were two dorsal fins, each with a spine, that of the 
first being the longer. There were no paired spines, 
and the ventral fin was opposite the second dorsal. 
The presence of an anal fin is doubtful. 4. The 
dentition was cladodont. 5. The vertebral axis was 
unsegmented, but there were extensive calcifications 
in connection with other parts of the skeleton.”’ 
It is obvious from this summary, that Cladodus was 
not at all related to Chlamydoselachus; and I may 
add, that it did not have the essential dentition of 
Chlamydoselachus, so well indicated by Mr. Garman 
SCIENCE. 
in the statement that ‘‘ each tooth has three slender, 
curved, inward - directed cusps, and a broad base... 
preventing reversion.” 
But, as Professor Cope has claimed (Science, vol. ili. 
p. 275), Chlamydoselachus did have a representative 
in the carboniferous genus Diplodus, or Didymodus; 
although I do not think that the two can be congener- 
ic. In fine, the recent discoveries by Messrs. Garman 
and Traquair enable us to co-ordinate a number of 
extinct types, and compel us, I think, to add two 
sub-orders or orders to the list of those necessary for 
the long-known living forms. The living sharks I 
have proposed (in Jordan and Gilbert’s Synopsis of 
the fishes of North America, p. 967) to distribute 
among four sub-orders; of which the Opistharthri 
or Notidanidae are the most generalized, and the 
Rhinae or Squatinidae the most specialized. The two 
additional sub-orders appear to be still more general- 
ized than the Notidanidae, and the sequence would 
therefore be-as follows: — 
1. Lipospondyli, including selachians without de- 
veloped vertebrae, but with a persistent notochord, 
and comprising the family Hybodontidae (Hybodus, 
Cladodus, Ctenacanthus, etc.). 
2. Pternodonta or Selachophichthyoidi, including 
Squali with vertebral condition unknown, and with 
teeth having fixed bases, comprising the family Chla- 
mydoselachidae (Chlamydoselachus and Didymodus). 
3. Opistharthri or Cyclospondyli. 
4, Proarthri (Heterodontidae). 
5. Anarthri (most living sharks). 
6. Rhinae. 
It is by no means certain that the hybodontids are 
Squali at all, and they may prove to be more nearly 
related to the Holocephali. The plate of Dr. Tra- 
quair’s memoir delineates very plainly one external 
branchial aperture, and one only; and the condition 
of the vertebral column and dorsal spines are features 
in which there is greater resemblance to the Holo- 
cephali than to the Plagiostomes. The primitive 
form from which the two diverged must theoretically 
have been not unlike the new Ctenacanthus, and it is 
quite possible that in the hybodonts we may have 
one of the ‘ missing links’ between the two groups. 
I had intended to refer to certain of the ‘sea-ser- 
pents’ which might be correlated with Chlamydose- 
lachus; such as the Maine animal noted recently in 
the Proc. U. S. nat. mus., the animal seen by Capt. 
Hope about 1848, and the selachian found in 1808, 
and partially described by Dr. Barclay, but must defer 
a notice to a future time. THEO. GILL. 
Evidence of unrecorded tornadoes. 
There is evidence in the forests of Pennsylvania 
that many localities have been visited by tornadoes of 
which no accounts have ever been recorded. The 
places referred to are called ‘windfalls;’ the timber 
having been prostrated apparently by violent storms 
of wind, while the trees immediately adjoining re- 
main erect and undisturbed. Sometimes, instead 
of forming a path through the forest, the tor- 
nado has descended, and quickly ascended into the 
air, leaving its marks on a small area, Judging by 
the remains of the timber-trees thrown down, these ~ 
events were of all ages, and of various degrees of 
violence. Sometimes the fallen timber was found 
sufficiently sound, after the first settlement of the 
country, to be worth manufacturing into lumber; in 
other cases, being older or more shattered, it was 
worthless: while in others it has entirely decayed and 
disappeared, the ground being covered with a later 
growth of a smaller and different kind, and the sur- 
