MORPHOLOGY OF THE VERTEBRATE OLFACTORY ORGAN. 339 



mouth must originally have been in front of them. Accord- 

 ing to Dr. Dohrn the present vertebrate mouth is formed 

 by the median coalescence of a pair of gill slits; my own 

 investigations lead me to the conclusion that, though these 

 gill slits do contribute to the formation of the mouth, there 

 is in addition a median involution of the epiblast of the 

 under surface of the head, as described by Balfour^ and 

 others, so that the mouth consists of three elements, a 

 median epiblastic involution and a pair of gill slits. I am 

 inclined also to believe that the fact of the olfactory organs 

 appearing in front of the mouth is due to two causes ; firstly, 

 the hypertrophy of the forepart of the head carrying the 

 olfactory sacs forwards ; and, secondly, an actual shifting 

 backwards of the median element of the mouth, of which I 

 think there is a certain amount of independent evidence. 

 The anterior end of the notochord is, as is well known, bent 

 completely round on itself, through an angle of fully 180^, 

 i. e. as Balfour has already noticed, to a much greater extent 

 than cranial flexure alone will account for. Now, assuming 

 that the notochord is a hypoblastic structure, and that its 

 anterior end remains for a time in connection with the hypo- 

 blast, a shrinking back of the hypoblast of the anterior end 

 of the foregut would at once account for this condition of 

 the notochord, and would at the same time cause a displace- 

 ment backwards of the mouth. It would appear therefore 

 quite possible that the median element of the present verte- 

 brate mouth is the original vertebrate mouth which has under- 

 gone a slight displacement backwards, and so has become 

 severed from the olfactory organs. 



Perhaps the most serious objection to the visceral-cleft 

 theory of the olfactory organ that is likely to occur at first 

 sight, is the fact that these organs are involutions of the 

 external epiblast, while the visceral clefts are formed by 

 diverticula of the hypoblast of the foregut. While fully 

 admitting the force of this objection, I venture to think that 

 the arguments I have brought forward — the evidence in 

 favour of the segmental value of the olfactory nerve, the 

 close relation, both anatomical and histological, between the 

 olfactory organ and the visceral clefts, the fact that these 

 relations are much more marked in the more primitive than in 

 the more specialised vertebrates, the various identities in time 

 of appearance and in histological structure, and the con- 

 current testimony of the various incidental circumstances to 

 which I have alluded — are sufficient to outweigh this objec- 

 tion. Moreover, we must bear in mind that slight ingrowths 

 I Op. cit., p. ISO. 



