242 Transactions of tJie Boyal Society of South Africa. 
" ethmoid just above the point where Meckel's cartilage articulates." It 
is of course clear that the tentacles cannot be mandibular gills if they 
are attached to the ethmoid ; but a consideration of the clause in inverted 
commas shows, I think, that the word "ethmoid" is a mistake and must 
be replaced by " palatal portion of the palato-ptery go-quadrate bar " (see 
my Fig. 7 and the discussion on the first visceral bar). If such a substi- 
tution is made Beddard's objection falls to the ground. But I have found 
that not only is the tentacular cartilage joined " above the point where 
Meckel's cartilage articulates," but also to the ethmoid in front of the 
nasal sac, i.e. to the cornu trabeculae ; and in view of this fact one must 
deny the homology of the Xenopus tentacles with the Triton balancers^ 
unless one accepts the theory that the trabeculae cranii represent a pair 
of premandibular visceral bars. Beddard is of opinion that the tentacles 
should be compared to those of a Siluroid fish or " perhaps better to the 
nasal barbels of Myxine and Bdellostoma." The barbels of the Siluroids 
are supported by the rudimentary maxillaries, i.e. in the embryo probably 
by Meckel's cartilage. In Myxine the barbels are supported by cartilages 
which are figured free. In Bdellostoma the barbel-cartilages are attached 
to the lingual bar (i.e. either basi-hyal or basi-branchial). The homologies 
of the tentacular bars, as expressed by Beddard, are rather vague. I am 
of opinion that they represent labial cartilages, even although Beddard 
found that the posterior root was formed as an outgrowth from the 
ethmoid (palatal) cartilage. 
Lungs. — Beddard observes that the high, columnar epithelium charac- 
teristic of the gill-slits does not extend into the glottis. My specimens, on 
the contrary, did show such an extension of the columnar epithelium, and 
therefore support the theory that the lungs and gill- slits are homologus 
organs (see Fig. 2, which is taken through the tadpole in a region 
posterior to the commencement of the groove leading to the third slit 
and still shows the tall epithelial cells medially ; these extend backwards 
into the glottis). 
Bespiration. — Sedgwick says that the larval respiration of Aglossa is 
entirely by means of lungs. Beddard states that since the filtering appa- 
ratus in the grooves leading to the gill-slits is the only observable respira- 
tory organs, they must perform the act of respiration. Marshall and 
BleSjf as quoted by Beddard, say, for the Frog, " as the blood is returned 
from them {i.e. filtering apparatus) to the somatic veins, it is probable 
that they are not actively respiratory." 
Beddard found external gills in the very young tadpoles, but did not 
know that they continued to be present at later stages, since he says that 
''they (filtering apparatus) clearly must be respiratory in Xenopus, as 
* Sedgwick, " Student's Textbook of Zoology," 1905. 
t Marshall and Bles, " Stud. Biol. Lab., Owens Coll.," ii., 1890. 
