250 FRANCIS DARWIN. 



contraction. This appearance may be due to a raoniliforra con- 

 traction in which the outlines of the beads and constrictions are not 

 perceptible, or may possibly be due to the contraction occurring at a 

 series of equidistant points, the refractive index being raised at 

 each point in consequence of the increased density of the tissue. 

 It should be added that contraction frequently occurs without 

 being preceded by moniliform appearance. 



I have hitherto spoken as if after the completion of the act of 

 contraction the filament were always found massed into a spherical 

 body on the gland from which it took its origin. This is nor- 

 mally the case, but in rare instances a long filament becomes 

 entangled with a neighbouring gland, and when contraction takes 

 place the filament snaps in two, one portion remaining attached 

 to its own gland and the remainder actually contracting on to the 

 foreign surface by which it had been entangled. It must be 

 noted that filaments frequently break loose from their attach- 

 ment, and while in this free state are capable of contracting and 

 exhibit other signs of vitality. These two observations are im- 

 portant because they throw some light on the nature of the act 

 which I have called contraction. A friend suggested that possibly 

 the filament may not undergo a true contraction, but that its 

 movements may be governed by changes occurring within the 

 gland, so that the filament itself is merely passively acted on by 

 these internal agencies. But it is evident that this cannot be the 

 case, because freely floating filaments are capable of contracting, 

 and because the distal portion of a filament exhibits the same 

 power independently of the proximal portion. 



Causes inducing contraction. — In the present section I shall 

 examine only those causes which produce the violent changes of 

 form which I have hitherto called " contraction.^' Under certain 

 conditions the filaments assume an amoeboid state in which 

 gradual alterations in shape spontaneously occur; it is obvious 

 that these movements may quite as justly be supposed to be due 

 to contractiUty ; but these movements are excluded from the 

 present discussion, and will be considered in the sequel. 



Acetic acid. — The contraction caused by this reagent has 

 already been described. The filaments are very sensitive to its 

 action, and contract violently when exposed to a solution of 1 

 per cent. Also distinctly with a much weaker solution, '2 per 

 cent. ; but on the other hand some filaments were found in a half 

 expanded state after three hours' immersion in '4 per cent, acetic 

 acid. 



After the protoplasm of the filament has been killed by the 

 action of acetic acid, a remarkable change occurs. The con- 

 tracted mass swells and becomes more transparent, and produces 

 ft pile of polyhedral vesicles, whose shape is determined by mutual 



