242 BULLETIN OF THE 
the egg of Agalma. Haeckel* says that he made attempts to fertilize 
artificially the ova of the genera Praya, Diphyes, Abyla, Hippopodius, 
Athorybia, Agalmopsis, Halistemma, Forskalia, Crystallodes, and Physo- 
phora. “Die Mehrzahl der Versuche schlug fehl, und in vielen Fallen 
gingen die befruchteten und sich entwickelnden Eier zu Grunde. Ehe 
sie noch iiber die ersten bereits von Gegenbaur beschriebenen Entwick- 
elungstudien hinaus die Entwickelungsvorginge zu verfolgen, gelang 
mir nur bei drei Physophoriden-Gattungen, namlich bei Physophora (bis 
zum XXVIII" Tage), Crystallodes (bis zam XXVII*" Tage), und bei 
Athorybia (bis zam VII*" Tage).” He does not state with sufficient ex- 
actness in the case of Crystallodes, the nearest ally of these three genera 
to Agalma, whether he artificially impregnated the ovum or not. 
The natural ovulation in Agalma was the only means of getting mate- 
rial for the study of the embryology ; and as this happened seldom, even 
in instances when I had in confinement a large number of large and 
sexually mature specimens, the amount of material at my control was 
small. The youngest larvee are very hardy, needing, for early stages at 
least, no change of water, provided decaying matter from the adult 
animal be not allowed to pollute it. 
Each ovum is carried in a bell-shaped structure called the female 
gonophore. The female gonophores (PI. I. fig. 1) are found in botryoidal 
bunches at the base of the polypites, and generally adhere to the neigh- 
boring hydrophyllium when it is broken from its attachment to the stem. 
_ The gonophore is fastened to the axis by its apex, through which a small 
tube communicates between a system of vessels called the radial tubes 
and the cavity of the stem. No marginal tube or marginal appendages 
of any kind were detected on the bell. The course of the radial tubes 
in the bell of the female gonophore (PI. I. fig. 3) is very irregular, and 
varies very greatly in different individual gonophores. In a form of 
gonophore which was common, the following arrangement in the disposi- 
tion of the tubes was observed. Two radial tubes arise from a common 
point under the apex of the bell, at the junction of the same with the 
tube of the apex. These lie in opposite hemispheres on the walls of the 
bell cavity. Consider the course of one of these radial tubes. After 
extending from under the apex of the bell about half-way down the 
sides of the bell on its inner surface, it bifurcates, each division passing 
at right angles to the course of the undivided tube. Each of the bifur- 
* Zur Entwickelungsgeschichte der Siphonophoren. Eine von der Utrechter 
Gesellschaft fiir Kunst und Wissenschaft gekroénte Preisschrift. Utrecht, 1869. 
p- 10. 
