248 
Transactions of the Boyal Society of South Africa. 
The structure of each opercular cavity may be described as follows : 
There is firstly an outer groove forroed by the upwardly bent opercular 
fold — this is the real opercular cavity, although only at the anterior end 
may it be called a cavity." Secondly, into this groove there open three 
pouches which I have called " external gill-pouches (Fig. 5, Gp 1 and 2). 
These three external gill-pouches may be imagined to have arisen by the 
partial fusion of the outer ends of three backwardly directed external 
gills — the last would of course fuse with the body-wall. Beddard actually 
saw external gills in very young tadpoles, and Fig. 5 shows two of these, 
Gl and G2. A section taken just in front of that represented by Fig. 5 
would show the first gill, Gl, attached ventrally, i.e. distally — it would be 
seen in the same way as is G2 in Fig. 5. The third gill commences further 
1 '~ — — ,'3 
Fig. 6. — 1 coelomic epithelium ; 3 hinder end of pronephros ; 2 nerve to arm ; 
5 connective tissue; 4 epidermis; 6 sac into which arm is invaginated. 
back than the level of Fig. 5 and is attached to the third visceral bar 
(B3). Each external gill-pouch is provided with a band of muscle fibres 
(two are shown in the two gills in Fig. 5) which runs around it and is 
attached to the branchial bars in front of and behind it. The contraction 
of these muscles would put the contents of the pouches under pressure and 
at the same time widen the opercular openings of the pouches. The wall 
of the external gill-pouches, i.e. the gills, are not vascular, and cannot 
therefore be actively respiratory. 
Fore-limbs. — The operculum, on each side, tails out just ventral to the 
end of the last branchial bar, while the fore-limbs are developed further 
backwards and, moreover, dorsal to the ends of these bars. The fore-limbs 
