28 NATURAL SCIENCE [July 
eulf which separates the invertebrates from the lowest vertebrates, 
the cyclostomes, can be bridged. I do not see how it is possible to 
demonstrate either the strength or the weakness of the particular 
bridge here sketched. It is enough if it is merely kept in mind as 
a hypothesis, that is by those to whom it appeals. Beyond sug- 
gesting this possible method of bringing about the more fundamental 
morphological changes which transformed the invertebrate into the 
vertebrate, the hypothesis does not profess to go; the subsequent 
development of skeleton, limbs, jaws, teeth, &c., belongs to vertebrate 
morphology as such. 
One possible test, viz., that of direct transition forms, is entirely 
lacking. If anything, however, this favours our argument. For 
if there be any truth in our suggestion we could hardly expect to 
find transition forms, at least certainly not fossilised, for such 
hirudinean ancestors would naturally be soft-bodied, and, by the 
time the notochord was developed in the manner sketched, they 
would be no longer hirudineans but primitive cyclostomes. 
One class of vertebrate, however, we do find, which is apparently 
lower than the cyclostomes—Amphioxus and the Tunicata. Though 
these have long been recognised as related to the vertebrates, their 
position in the chain of evolution has been matter of endless con- 
troversy, for the obvious reason that the order of the chain is not 
known. It seems to me that the line of development from the in- 
vertebrate hirudinean to the vertebrate cyclostome which is here 
suggested would assign them a natural position. They are not true 
links in the ascending chain, but forms which have branched off and 
become differentiated, again, be it specially noted,in adaptation 
to a new method of feeding. We postulated for our earliest 
ancestors when they were just turning into vertebrates, an active 
free-swimming life, catching and devouring food with gaping jaws, 
and consequently with a stream of water flowing through the 
pharynx and out at the gill-slits, kept up by rapid swimming 
through the water. This constant stream of water through the 
pharynx would carry along with it innumerable small particles of 
food, which might with advantage be caught and turned down the 
oesophagus. It seems probable that all the early vertebrates de- 
veloped an organ for this purpose, the endostyle, the particles 
perhaps attracted by and adhering to a slimy secretion which was 
then worked by ciliary motion into the oesophagus. Now, either in 
special regions where these small particles were unusually plentiful, 
while larger prey requiring great exertions and the use of teeth was 
scanty, or among very young forms not yet strong enough to attack 
prey with mouth and teeth this passive method of feeding was 
nearly sure to be adopted by a certain number of individuals as 
1 The larval amphibian has acquired a browsing herbivorous diet. 
