188 FOOT-NOTES TO EVOLUTION. 
istics of both. Thus the hybrid presents a new type, 
which was found to be likewise constant, and in no 
case did it approach the likeness of either the Zchinus 
or Spherechinus type, so as to be mistaken for them. 
Thus were established four preliminary essentials to 
the experiment, viz.: 
1. The constancy of the normal Zchinus larval type. 
2. The constancy of the normal Sphe@rechinus larval 
type. 
3. The constancy of the new hybrid type between 
Echinus & and Spherechinus 9 . 
4. The possibility of securing non-nucleated frag- 
ments of Spherechinus eggs by shaking, which were 
capable of fertilization, and from which dwarf larve 
might be reared. 
The crucial point in the experiment, as may have 
been already anticipated, was to cross-fertilize non-nu- 
cleated fragments of Spherechinus eggs with Echinus 
sperm. The type of larva resulting would decide the 
point in question. Here we may have a nucleus from 
one species introduced into the non-nucleated cyto- 
plasm of another species by a perfectly normal process 
—i. e., by the penetration of the spermatozoan into the 
egg. If the resulting larva be of the hybrid type, as 
shown in Fig. 14, A and B, then nucleus and cytoplasm 
both unite in determining the hereditary characteristics ; 
if the pure Spherechinus type of Fig. 13, Cand J, results, 
then the cytoplasm alone bears these influences; while, 
finally, if the type be that of Fig. 13, 4 and B, to the 
spermatozoan nucleus alone must be ascribed the heredi- 
tary qualities. The larve secured by Boveri were of 
the pure Echinus or paternal type, as is well shown in 
Fig. 14, Cand D. A comparison of these figures with 
Fig. 13, 4 and £&, show unmistakably that this is the 
case. No trace of the Spherechinus, or maternal, char- 
