POLYCHEHTA FROM THE N.E. PACIFIC. 979 
as the result of an unexpected stimulus*; (2) that, following 
autotomy, regeneration will take place from the posterior fragment, 
the median region thus giving rise to the whole of the anterior 
region, and after that replacing the missing median segment. 
There is thus a clear resemblance to the manner of regeneration 
in Phyllochetopterus and also a clear minor distinction, the 
presence of a definite breaking-point in Chetopterus and its 
absence in Phyllochetopterus. Autotomy and a complete type of 
regeneration are thus to some extent characteristic of the family 
Cheetopteride. In Phyllochetopterus prolifica, however, the occa- 
sions on which autotomy takes place are so frequent and regular as 
to subserve a definite method of asexual reproduction. But the 
nature of the stimuli which cause autotomy, and the question 
whether the phenomenon is in any sense under the control of the 
animal itself, can hardly be approached as yet. 
Some advantage may, I think, be gained by comparing the 
cases of regeneration studied in other Polycheta with a differentia- 
tion of regions. Ivanow (8) and other authors have made a very 
thorough examination of these phenomena in the case of the 
Sabellid Spirographis spallanzani. Here there are three regions : 
the anterior thoracic with the prostomium, bearing the enormously 
developed tentacles, and the first three setigerous segments; the 
posterior thoracic, consisting of eight or nine following segments ; 
and the abdominal, with an indefinite number of segments. Only 
such fragments regenerate as consist of abdominal segments or of 
abdominal and thoracic segments. ‘Those containing thoracic 
segments only always disintegrate. In regenerating fragments, 
the hinder end always produces abdominal segments, and the 
anterior end regenerates the prostomium and the three anterior 
thoracic segments. The posterior thorax develops later by the 
metamorphosis of the most anterior abdominal segments, a 
striking change taking place in the characters of the parapodia. 
The dorsal uncini are replaced by capillary sete, while in the 
neuropodium the capillary setz are replaced by uncini. 
A similar phenomenon has been described by Watson (12) in 
Potamilla reniformis, another Sabellid. Here, in the regeneration 
of the anterior region from abdominal fragments, the prostomium 
and one new setigerous thoracic segment only are formed as a new 
growth: all the rest of the thoracic segments are formed from 
abdominal segments in which a modification of the parapodia 
like that described above occurs. It is curious that two regions, 
differing from each other so little in morphological characters as 
do the anterior and posterior thoracic regions of Spirographis, 
should have such a dissimilar method of re-formation. 
Though my observations on regeneration in Phyllochetopterus 
* In 1913, at Plymouth, I noticed that of a tubful of Chetopterus which had 
been brought i in, after being kept on board a trawler for 20 hours or so, nearly all had 
autotomised, as a result of “the unhealthy conditions, rupture taking’ place between 
the first and second segments of the median region. 
66% 
