522 TRANSACTIONS OF SECTION D. 
and functional organ. It is wanting in many rodents, some insectivores, and 
most bats. 
Hammar derives the mammalian palatine tonsil from the dorsal diverticulum 
of the second gill-pouch. This would appear to hold good for Eutharia at all 
events, though Griinwald derives the human tonsil from the ventral diverticulum. 
Hammar divides tonsils into (a) primary and (b) secondary, according as to 
whether lymphoid tissue is laid down in the tuberculum tonsillare or not, placing 
man and the artiodactyla in the second division, and most other mammals in 
the first. This classification is at best somewhat arbitrary and presents several 
difficulties, e.g. :— 
1. It is difficult to classify all tonsils on these lines, as there are numerous 
intermediate forms. 
2. It is difficult to interpret the tonsils of primates other than man on this 
basis. 
3. Although in man lymphoid tissue in the plica triangularis (=the tuber- 
culum tonsillare) is secondary, yet the hypertrophy follows definitely anatomical 
lines, so that such tonsils may be classified under several distinct types. 
Investigation of a large number of mammalian tonsils shows that the gross 
anatomy is very distinctive in each group, being always characteristic of the 
order, and frequently also of the family, or even, in some cases, of the genus. 
It is not easy to show an actual correlation between structure and habit in the 
case of this organ, but it is worthy of note that the carnivorous Marsupials have 
tonsils which bear a remarkable resemblance to those of Eutherian Carnivores. 
MONDAY, SEPTEMBER 15. 
The following Papers were read :— 
1. Pseudo-hermaphrodite Examples of Daphnia pulex. 
By Dr. J. H. Asuworru. 
2. Evolution of the Caudal Fin of Fishes. 
By R. H. Wurrenovuse, M.Sc. 
The diphycercal caudal fin of fishes may be primitively or secondarily sym- 
metrical; the former is the protocercal type, and the latter the gephyrocercal. 
The diphycercal palzozoic fishes are usually considered gephyrocercal, but it is 
possible that they are survivors of the original protocercal form. 
In the heterocercal form—e.g. Heterodontus—the anterior dorsal caudal 
radials are well removed from the neural arches, but they gradually approximate, 
and finally fuse together as the posterior end is approached. The same con- 
dition of things once occurred ventrally, but with the greater development of the 
ventral lobe, and the consequent necessity for a firmer support for the fin rays; 
the fusion took place further forward, and the gradual approach of radials to 
hemal arches is not shown. Still, evidences of such fusion are frequently seen, 
and the hypurals of the fin are shown to be the result of a fusion of radial and 
hzemal arch. 
Such evidences as these are interesting in connection with problems concern- 
ing the morphology of the various elements in the homocercal fin, which is 
nothing more than a specialised heterocercal form. 
The urostyle, which is a fusion of a number of vertebre, may be present as a 
slender rodlike bone, or it may be entirely suppressed. Owing to the over- 
crowding due to the excessive upturning of the caudal extremity, the epaxial 
arches of the urostyle are usually altogether suppressed; on the other hand, 
ventrally they remain and become greatly developed, owing to increased oppor- 
tunities resulting from the upturning. 
In the great majority of homocercal caudal fins dorsal fin-rays are supported 
by radials; this is interesting, as it shows that radials are more persistent than 
arches when subjected to adverse conditions, such as overcrowding. 
