4 Maciteay Memortat Vouume. 
Haswell was good enough to make this copy for me, the plates were missing. If I 
fail to acknowledge work previously done it is simply owing to inability to refer to 
this in Melbourne. Judging, however, from the papers by Giinther,* Boas and 
recent text books, little has been described beyond the structure of the heart and the 
relation of some of the main vessels, and it appeared to be worth while to describe as 
fully as possible the whole circulatory system of Ceratodus at the risk of traversing 
ground already passed over to a certain extent by previous workers. 
IJ.—Tue Arrerizs. 
A. The ventral aorta and afferent branchial arteries. 
The ventral aorta in Ceratodus, as in other Dipnoi, is extremely short ; in fact, 
the four afferent branchial arteries, as shown by Boas and Giinther in Ceratodus and 
by Owen and others in Lepidosiren, arise almost directly from the anterior end of the 
strongly muscular conus arteriosus, their roots being close together as in the 
Amphibia. 
When lying in their normal position (fig. 1), the first and second are nearest to 
the ventral surface of the body; slightly dorsad of these the third enters and covered 
over by the third the fourth enters. In the normal position, the third artery must 
be pulled aside before the root of the fourth can be seen. 
The arteries pass forwards and at the same time bend outwards and slightly 
dorsally to enter the four branchial arches (fig. 20), each of which bears a holo- 
branch.+ 
The gills have been already described and figured by Giinther{; but as there are 
one or two points of importance connected with the relationship of the blood vessels 
dependent upon the exact distribution of the branchial lamine, the gills are here 
described and represented in figs. 20 and 21. With minor exceptions the following 
account agrees closely with that given by Giinther. The anterior hemibranch is 
attached to the hyoidean arch, and may hence be conveniently termed the hyoidean 
hemibranch. This is equivalent to the opercular gill of such forms as Elasmobranchs, 
Holocephali, Chondrostean Ganoids, and Lepidosteus, and is spoken of by Giinther 
as the pseudobranchia. 
* Description of Ceratodus, a Genus of Ganoid Fishes, recently discovered in Rivers of Queensland, Australia, <A. 
Ginther. Phil. Trans. Roy. Soc. London, 1871, p. 511. 
+ In the use of this convenient term, as in that of hemibranch, I have followed Prof. T. J. Parker. A hemibranch 
is the whole set of gill filaments on one side of a gill arch. A holobranch is composed of the two sets, one on either side of 
a gill arch. 
+ Loc. cit. p. 539, pl. xxxvu. fig. 6. 
