Organism of the Siphonophora. 193 
genetic development of the Medusa by Hydroid-stocks entirely 
out of consideration. That is the central and at the same 
time the weakest point of the theory, which at once brings 
the opposition to the Hydroid-theory into prominence. This 
commences with the older and original metagenetic develop- 
ment of the stem-form, and refers the resemblance to a bilate- 
rally constructed Medusa which makes its appearance so early 
in the young Siphonophoran larva, only to external analogies 
secondarily produced. In this the Siphonophoran larva does 
not appear as the repetition of a primitive, hypogenetically 
reproducing, bilateral Oceanid with dislocated stomachal tube 
and marginal filaments, which by continued gemmation of 
new Meduse and parts of Medusz produces the polymorphic 
stock, but a free-swimming developmental stage of the 
Hydroid-stock of an Oceanid reproducing metagenetically, 
furnished the starting-point for the production of the Siphono- 
phora, and in fact the prevention of fixation was the cause of 
the first change, the occasion of a series of transformations 
which then also affected the sexual Meduse budding forth 
from the stock. Of course, in the absence of any data fur- 
nished by transitional stages and intermediate forms, it must 
be left to fancy to finish the picture of the changes through 
which in the phylogenetic process the original form resembling 
a larval Hydractinia or Podocoryne could have transformed 
itself into a Siphonophoran. It is only in this light that the 
attempt made in my little paper is to be judged, as a repre- 
sentation which, when compared with the picture of the 
budding Medusa, has at least an equal justification. The 
reconciliation between the Medusa- and Hydroid-theories 
which I attempted in this statement therefore depended upon 
the proof that, while for the former the conception of the 
Siphonophore as a polymorphic stock appears by no means 
excluded, the second theory also presupposes the presence in 
the stem-form of a Hydroid Medusa, I could approve of the 
Medusa-theory in so far as it starts from the Hydromedusa, 
but could not concede to it that this is to be found repeated 
even in the primary larva, and that the latter was to be palin- 
genetically interpreted. On the other hand, I defended the 
Hydroid-theory, in the conception of the polymorphic stock 
in which I found no contradiction to the tormer with reference 
to the starting-point of the Siphonophora, which is to be 
sought not in the mature Hydroid-stock, but in the free- 
swimming larval stock. The supposed stem-form was not a 
symmetrical Medusa with dislocated organs and hypogenetic 
development, but a metagenetically developing, normally con- 
structed Medusa, in the swimming larval stocks of which the 
