SOME PROBLEMS OF COELENTERATE ONTOGENY 541 
spheres, spindles, etc., are held to be types of normal nuclei, I 
think it must be allowed that the burden of proof that the former 
are but phases of the latter is upon the champions of exclusive 
mitosis, and thus far the evidence which they submit has not been 
convincing. 
With the facts herein presented, and the cumulative evidence 
available from a wide range of observation and authority clearly 
appreciated, it seems difficult to evade the force of the conclusion 
which is implied. The writer believes, therefore, that his earlier 
tentative suggestions concerning amitosis as a mode of nuclear 
activity is not only not discredited nor disproved by later 
researches, but is rendered both credible and probable. 
SUMMARY 
The main points embodied in the paper may be summarised 
as follows: 
1. Later observations on the development of Pennaria, and 
including a new species, go clearly to confirm the earlier results, 
and to show that it is not peculiar to a single species or to a given 
locality. 
2. Observations on the development of Hydractinia echinata 
also confirm much of that found to occur in Pennaria, including 
cleavage, ectosarcal features, formation of germ layers, etc. 
3. The same may be said in general as to Clava leptostyla. 
New facts as to certain histogenic aspects seem established, and 
the significance of the early embryo,—the morula,—is empha- 
sized. 
4. Concerning the origin and growth of germ cells it becomes 
more and more certain that the theoretical contentions of Weis- 
mann find no warrant in the ontogeny of Coelenterates, and par- 
ticularly in that of Hydrozoa, the group especially claimed by 
him. 
5. A review of earlier doctrines of homology goes to show that 
they have been greatly overestimated as criteria of phylogeny. 
This includes especially the features involved in germ layers, the 
early hydroid larva, cleavage homology, prelocalization of ger- 
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3 
