PRELIMINARY ACCOUNT OF DEVELOPMENT OF LAMPREYS, 151 
immediately above the mouth; then the epiblast cells which 
border the pit above are thickened and form a layer of epithe- 
lium which is perfectly continuous, and takes up the whole of the 
anterior part of the head. On account of the cranial flexure 
this epithelium looks directly downwards. The pit becomes 
gradually deepened, but the olfactory epithelium remains still 
on the surface of the head; soon, however, it has a deeper posi- 
tion, and shows only a small triangular opening on the exterior. 
Late in larval life the epithelium of the olfactory organ develops 
the well-known folds which show a definite paired arrangement. 
The palato-nasal passage does not attain any considerable length 
until after the metamorphosis; and its rudiment is from the first 
single. The paired olfactory nerves show, however, that that 
organ was at one time paired, and that a later fusion of the two 
pits took place. If the two pits of the Sturgeon (O. Salensky, 
]. c., Taf. 1x, fig. 84), which lie at the anterior part of the head 
and not laterally as in the Elasmobranchs, were brought nearer 
the middle line, we would have almost exactly the condition of 
Petromyzon. But in the latter type the paired stage has been 
over-lept in the course of development. 
The mesoblast, in the earlier stages, has in general very much 
the relations that are observed in the Elasmobranchs, or still 
more like those of Triton ; but it is worthy of notice that in 
Petromyzon the first pair of protovertebre follows closely upon 
the auditory vesicles instead of leaving a considerable interval 
free between them as in the Hlasmobranchs. The protovertebre 
develop muscle-plates which gradually grow forwards and overlap 
each other; the anterior ones grow over the head as far as the 
olfactory capsule, and by processes of division develop the 
muscles of the head. But since these myotomes belong mor- 
phologically to the trunk they cannot be regarded as guides to 
the segmentation of the head. 
In the head the mesoderm undergoes another segmentation, 
giving rise to one segment between every two gill-slits and two 
in front of the first gill-slit, just as in the Elasmobranchs and 
Urodeles. These segments surround a central cavity and cor- 
respond exactly to what Balfour has called “ head-cavities.”’ 
These segments develop the gill muscles, and I believe the first 
pair give rise to the muscles of the eye. In the trunk the 
development of the mesoblast, its splitting formation of the 
pleuro-peritoneal cavity, &c., shows no essential deviations from 
the conditions found in the Elasmobranchs and Urodeles. The 
internal muscles of the sucking-disc appear to develop them- 
selves directly from indifferent mesoblast cells. The formation 
of the sucking-disc is very striking, and its peculiarities appear 
very early. At first the upper lip appears as a low rounded 
