EMBRYONIC DEVELOPMENT OF THE TELEOSTEANS. 45 



granular, but devoid of any globule. I took the very greatest 

 pains to ascertain whether there exists in this layer any structure 

 which might be considered as a cell-nucleus. Neither on exa- 

 mination of the fresh, living eggs, nor any more after having 

 treated them with ostnic acid and colouring matters (picro- 

 carmine, Beale^s carmine, hsematoxylin), did I find any trace of 

 anything of the kind. Acetic acid of 1 per cent, dilution gave 

 me no better results. This layer forms for the deutoplasmic 

 globe a continuous investment of such a sort that at no point do 

 the segmentation spheres rest directly on the deutoplasm. It 

 extends in every direction beyond the margin of the segmented 

 disc, and everywhere is seen to be closely spread upon the 

 vitellus. It is not easy to see the limit of this protoplasmic 

 m.antle, so fine does it become at its margin. It does not present 

 the same thickness in every part. At the centre beneath the 

 segmented germ it forms a biconvex lenticular body which 

 occupies all the space between the vitelline globe, which • is 

 depressed at this point and the deep face of the segmentation 

 spheres. In addition it presents a considerably increased thick- 

 ness beneath the margins of the germ. This thickening forms 

 all round the germ a circular walJ, triangular in section. One 

 side of the triangle is adjacent to a segmentation sphere ; it 

 faces upwards and inwards towards the animal pole of the egg ; 

 the second side is adherent to the nutritive vitellus, and concave. 

 It faces downwards towards the vegetative pole of the egg. The 

 third side is free ; it faces directly outwards, and is slightly 

 convex. It is separated from the membrane of the egg by a 

 sj)ace filled with a clear and hyaline liquid. This layer is homo- 

 logous, as I shall show further on, with that which Van Bam- 

 beke has described in the Roach by the name of intermediate 

 layer. This name is appropriate on account of its interposition 

 between the germ and the deutoplasmic globe, and has the great 

 advantage of not in any way prejudging its morphological value. 

 I shall then designate it by this name : I shall call the thickening 

 which occurs at its centre beneath the germ the median lens 

 (lentille) ; with Van Bambeke I shall call the circular thickening 

 subjacent to the margins of the segmentation disc the lieripheral 

 welt (bourrelet). 



I was able to see the successive phases of the segmentaHon by 

 examining at small intervals new eggs removed from the mass 

 which was brought to me at seven o'clock in the morning, and 

 in which the eggs were at that hour provided with two seg- 

 mentation spheres. About half-past eight all the eggs exhibited 

 the disc segmented into four ; so that at least an hour and a 

 half, and probably more, must elapse between the appearance of 

 the first groove and the moment when the two first spheres 



