522 Piercy . — The Structure and Mode of Life of a 
straight filaments, and were then noted to occur usually between cells which 
contracted at different rates on loss of water ; in these cases, too, they 
affected the entire circumference of the septum, without extending right 
across it. As the lamellae of a septum were gradually separated by the 
breaking down of the thickening substance between them, they became 
bulged in a convex manner as is usual with free end walls (Fig. 3, L). 
Attempts to produce splittings by wetting dry filaments under the 
microscope, and to extend partial splittings by successively drying and 
wetting the filaments, were unsuccessful. The latter treatment, nevertheless, 
caused a swinging movement in curved filaments, and probably, where the 
rate of contraction was different in adjacent cells, a definite strain on the 
cuticle at the transverse walls. Dry filaments were, however, observed to 
shear at septa adjacent to dead cells, on becoming suddenly turgescent. 
In a consideration of the causes of splitting, the ‘ cuticle ’ may be 
treated as a thin cylindrical shell subject to internal fluid pressure. This 
assumption appears to be warranted by the observation that cells separated 
by a split often widen at their free extremities (Fig. 3, m), where the 
c cuticular ’ sheath has given way, this indicating that the latter constitutes 
a distinct strengthening layer of the wall. The ‘ cuticle 5 should therefore 
be especially capable of resisting a transverse tear. 1 When a filament is 
bent, a greater longitudinal stress is set up in the ‘ cuticle 5 at the outside of 
the curve, but splitting occurs also in straight filaments and then begins 
round the whole circumference of the filament. It would therefore seem 
that splitting at a septum, if occasioned by increase of turgor as described 
by Klebs, can only arise where there is a general or local weakening of the 
c cuticle ’ at the septa. It has been shown above that a more or less general 
degeneration of the ‘ cuticle ’, frequently pronounced at the septa, does 
occur in filaments which are in process of splitting. Such a degeneration 
probably results from the folding and stretching of the ‘cuticle’ caused 
by curvature of the filaments, and from the separation of the ‘ cuticle ’ 
from the longitudinal wall, occasionally observed when a filament contracts 
owing to loss of water (Fig. 3, N). One reason for its prevalence in the 
regions of the septa may be the difference in the rate of contraction and 
possibly expansion, occurring among individual cells of the same filament, 
during changes in turgescence. On the other hand, splitting might result 
from a distinct longitudinal force being brought into play between the two 
lamellae of the septum, tending to separate them. A special weakening of 
the ‘ cuticle ’ adjacent to the septa must be produced by the development of 
the mucilaginous substance, and the continued extension of this substance 
possibly produces a longitudinal force which ultimately ruptures the 
‘ cuticle 
1 The resistance of such a cylindrical shell to a transverse tear is double its resistance to 
a longitudinal tear. Cf. D. A. Low : Applied Mechanics, p. 76. 
