122 REVUE BRYOLOGIQUE 
This explanation was however found inadequate to explain 
certain of the cases, notably such as shown at (d), where only a 
very short portion of the seta has been split off, while the capsule ; 
itself is deeply ruptured. This is a case of frequent occurrence. 
Any search for the origin in the attacks of insects is out of 
the question; there is no indication of anything of the kind, à 
and moreover in no case is there any loss of tissue, the semi-de- 
tached portion always fitting precisely into the groove or wound 
left on the seta. 
I am inclined to attribute the abnormality to some disturbance 
in the physics of growth, associated with a tendency to abnprmal 
Structure of the seta. À considerable number of setæ show in 
varying portions of their length a tendency to flattening out, 
which in some cases is so pronounced as to render it almost ribbon 
shaped, the longer diameter being perhaps three times the thick- 
ness of the shorter. I have figured an extreme case of this kind 
- at fe). Here the upper portion of the seta is flattened out, witha 
longitudinal median groove, which even passes up into the capsule, 
“while at a certain point the seta is actually split in two, reuniting 
below, and a little lower down becoming entirely normal. If the 
splitting had been carried a little further, both upwards and 
downwards, the result would have been very similar to the usual 
= form as shown at fa) and {b). This monstrosity of seta is not 
_ however necessary to the production of the abnormality, as 
in many cases the split seta appears to be absolutely normal 
in structure ;the capsule base however in. most of the abnormal 
Cases seems to be somewhat gibbous instead of gradually tape- 
- ring into the seta as it should do in this species. This however is 
far from being universal, and in many of the best marked 
examples of the abnormality it is impossible to detect anything 
_ unusual in the original form and structure of either the capsule 
or the seta. 
= The most probable cause of the tearing away of the strip of 
_ Lissue would seem to be sought in the torsion of the seta. À strong 
 twisting of the seta, aided by a predisposition to longitudinal 
_ fissure by an abnormal structure of the tissue might quite concei- 
_ vably produce a splitting which would ultimately tear off the * 
Strip at its lower end. This suggestion is supported by the fact 
: on in all the examples which 1 have moistened out the semi= 
+ with the lens. But rather curio 
