180 REPORT—1905. 
mesoderm are differentiated in its roof, but situated wholly within the 
animal hemisphere. 
(8) These embryos resemble the last, but neither medullary groove, 
notochord, nor mesoderm is formed. 
(9) There is, perhaps, a slight blastoporic invagination, but the animal 
hemisphere is so exceedingly rough and corrugated that this is hard to 
make out. 
Part of the roof of the segmentation cavity is thickened and infolded ; 
this seems to represent a medullary plate. 
Urea.—The original concentration of urea employed was 1-14 per cent. 
The eggs were placed in the following more concentrated solutions : 
(1) 1:17 per cent. ; (2) 1°33 per cent. ; (3) 1:56 per cent. ; (4) 1:87 per 
cent. ; (5) 2°34 per cent. 
a. 2°34 per cent.—The yolk is nucleated, but unsegmented ; the animal 
hemisphere is incompletely segmented ; it consists of a mass of small uni- 
or multi-nucleate masses covered by a layer of flat cells. The nuclei are 
either very large, pale and reticular, or small and homogeneous. Very 
often the animal hemisphere is thrown into round, projecting, multi- 
nucleate masses, which recall the ‘framboisia’ of Roux. The egg does not 
advance beyond this condition. 
B. 1:87 per cent.—In its effects this solution closely resembles the last ; 
there are, however, more nuclei in the yolk. 
y. 1:56 per cent.—The contrast between this and the last is rather 
remarkable. The medullary folds are wholly closed ; there is, however, 
a, small persistent yolk-plug ; the tail is short and provided with a fin ; 
there is a nostril, and the external gills and gill-slits are formed. A bent 
heart, fairly large pericardium, blood-vessels, and auditory vesicle are 
formed ; the pronephros has only one funnel, and the duct is not open to 
the cloaca ; the retina is well developed, with the choroid fissure, but 
there is no lens. 
Quite the most remarkable feature of these embryos is, however, the 
doubling or tripling of the notochord. 
In addition to the usual vacuolated chord between the roof of the 
gut and the spinal cord, two other notochords, as they must be termed, are 
found, sometimes three others : (1) The cells which form the roof of the 
gut take on the characteristic vacuolation of notochordal tissue ; the mass 
of cells so formed is irregular, but certain parts of it become rounded off 
and assume a cylindrical shape like the notochord proper. This ‘enteric’ 
notochord, as I will call it, is fused here and there with the notochord 
proper, and is found usually about the region of the heart only. 
(2) A certain lateral or latero-ventral portion of the wall of the 
medullary tube behaves in precisely the same way, the vacuolated cells 
actually lying next the lumen, This ‘neural notochord’ may, but does 
not necessarily, fuse with the notochord proper ; it, too, does not extend 
the whole length of the body. 
(3) Lateral ‘mesodermal notochords’ may be formed in the same way. 
This occurrence of a specific kind of histological differentiation in an 
abnormal situation is a point of considerable interest, showing, as it does, 
that the prospective potentiality of gut roof, protovertebre and medullary 
tube is not yet fixed. It also suggests the possibility that the formation 
of the normal notochord may be due to the accumulation—in the mid- 
dorsal line—of the excretory products of the extremely active metabolism 
