ON PROTRUSION OF PROTOPLASMIC FILAMENTS. 253 



drop or two of mettylated spirit. The subsequent effect of 

 alcohol has puzzled and astonished me more than any other point 

 in this research. First of all, the contracted mass seated on the 

 top of the gland turns of a greenish colour. It then begins to 

 diminish in size, and ultimately either almost or quite disappears. 

 At first I was repeatedly deceived by this appearance, and believed 

 that the filament was actually retracted within the gland. What 

 really happens is that a great part of the filament is dissolved by 

 the alcohol. The reaction with alkanet shows that resin is con- 

 tained in the filaments. I presume that it is spread through the 

 protoplasm of the filament, and corresponds to the metaplasm of 

 Hanstein, or is intimately connected with the protoplasm in some 

 other way. The results of treatment with alcohol seem to show 

 that the quantity of resin is very large compared with that of 

 the protoplasm. In some cases a minute shrivelled ball remains 

 after treatment with alcohol ; in other cases the contracted fila- 

 ment breaks loose and floats away before the alcohol has com- 

 pleted the solution of the resin contained in it ; in a third set of 

 cases the whole of the contracted filament disappears under the 

 influence of alcohol. I know not how to explain this phenomenon. 

 The summit of the gland is sometimes hollowed out slightly, and 

 is difficult to examine accurately with high powers ; it is possible 

 that the minute remnant of protoplasm remaining after the resin 

 is dissolved, and which would necessarily be shrunk by the 

 alcohol, might be overlooked within the hollow on the summit of 

 the gland. Again, it is conceivable that if the resin is very 

 intimately distributed throughout the substance of the filament, 

 its sudden removal by a powerful solvent might cause the disin- 

 tegration of the remnant of protoplasm. 



In a filament killed by osmic acid in an extended condition, 

 and which was certainly dead (as it did not contract with 20 per 

 cent, acetic acid), an effect of alcohol was seen which I cannot 

 explain. The specimen was irrigated with methylated spirit with 

 the intention of dissolving the resin of the filament and leaving a 

 protoplasmic skeleton in an extended position. But the filament 

 ran together in a manner which could not be verbally distin- 

 guished from contraction, although it had more the appearance of 

 a filament of spun glass melting into a button than the normal 

 act of contraction. 



Chloroform. — Applied in the form of a vapour, choroform 

 causes contraction. A thin transverse section of a young leaf 

 was suspended by means of a drop of water to the under surface 

 of a thin glass cover, forming the roof of a gas chamber. The 

 chamber had the usual arrangement of tubes, one being connected 

 with a washing bottle (in which the chloroform is placed, covered 

 by a layer of water), the other either with the mouth of the ob- 



