Chap. VII. LAOMEDEA AMPHORA. 313 



rising together and constricting at their base, form a globular sac with a short and 

 moderately thick peduncle. In the mean while, the space between the walls has 

 gradually increased in size, but is constantly filled by the egg, which develops 

 at the same rate, until the medusa has matured {Fig. IG), when the egg occupies 

 about four fifths of the whole bulk of the projecting body. In this state the 

 inner wall {Fig. 16, h) is inverted upon itself, and constitutes a shallow, saucer- 

 shaped basis for the egg. The chymiferous cavity {h*) penetrates to the extreme 

 edge of the saucer, where, in profile, it appears like an incipient stage of the 

 radiating chymiferous system. In later stages, when the segmentation of the yolk 

 is going on, the saucer gradually diminishes, and finally becomes a mere disk 

 (PI. XXXI. Fiffs. 5, 5", 6, and 6", h) or truncate termination {Fiys. 7 and 8, h) of 

 the interior wall, and, at the same time, the yolk mass gradually fills the space 

 left by the retreating saucer, and, finally, becomes a globular mass {Fig. 8, ae). 



The egg (PI. XXXI. Fig. 2, ae) is always more or less flattened, even at maturity 

 (PI. XXX. Fig. 16, ae), when upon the point of undergoing segmentation. It con- 

 sists of a dense mass of minute yolk granules (PI. XXXI. Fig. 2, ae), and a large, 

 tough, clear Purkinjean vesicle {Figs. 2, jh and 2"), which contains several irregular, 

 scattered mesoblasts, and, within each of the latter, one, two, or three very minute 

 granular entoblasts. The process of segmentation is very easily traced, on account 

 of the moderate degree of opacity of the yolk ; it commences by forming a 

 furrow (PI. XXXI. Figs. 3, 3% and 3", a a} c?) across the yolk on that side which 

 lies next to the peduncle of the medusa, that is, on the abactinal side. The 

 division proceeds very rapidly ; in fact, it could actually be seen, for, in one hour, 

 not only had the yolk separated into two, but each half had divided again into two 

 {Fig. ^, b c d e), by furrowing transversely (/ g) to the 2:»rimary constriction {a a^). 

 That the segment masses are not always of equal size among themselves, may be 

 seen in two of our figures {Figs. 5 and 5"), which were drawn carefully to illus- 

 trate this point, and lettered correspondingly with Fig. i. It will be noticed that 

 the first and second constrictions {Fig. 4, a, a}-f, g) pass through the yolk in planes 

 which are j^arallel to the axis of the medusa, but at right angles to each other. 

 In the next stage, each of the four segments divides in a direction either directly 

 {Fig. 6, h c, b^ c') or obliquely transverse {Fig. 6", d e, d^ e^) to the axis of the 

 medusa, so as to form eight segments. As the self-division goes on, the yolk 

 gradually becomes less opaque, so that, by the time it is separated into thirty- 

 two masses {Fig. 7, ae af ag), the granular contents {Fig. 7°^) of each segment may 

 be seen without difficulty. Here, too, as in former stages, the segment masses 

 vary considerably in size ; some of them {ag) being fully one third greater in 

 diameter than others {ae). It i.s, also, a yei-y notable fact that the yolk, as a whole, 

 diminishes in bulk, as segmentation proceeds (compare Figs. 6, 7, and 8) and the 



VOL. IV. 40 



