ON PROTRUSION OF PROTOPLASMIC FILAMENTS. 247 



formed in drops in a vacuolated protoplasm in the gland-cells of 

 the trichomes, and, passing through the cell-walls, collects on 

 the external surface of the gland. In his paper on glandular 

 trichomes Hanstein' describes a method of secretion ia which 

 the product collects in a blister-like cavity beneath the cuticle, 

 which bursts and allows it to escape. On dissolving the resinous 

 crusts from the teasel glands I have occasionally seen a similar 

 bladder of cuticle subsiding or bursting. Martinet^ describes a 

 similar process, and considers it pathological and of only 

 occasional occurrence. I imagine that in the case of the teasel 

 the essential act of secretion, i.e. the elimination of a resinous 

 product from the cells in which it is formed, is a normal 

 process ; but possibly the method of elimination may vary, so 

 that a bladder may or may not be formed. The resinous matter 

 is secreted only by the trichomes of the types shown in figs. 2, 

 3, 4, 13, &c., that is to say, only by those from which the 

 motile filaments are produced. 



T/ie motile filaments. — Various forms assumed by these 

 remarkable bodies are shown in figs. 2, 3, 6, 7, 8, 9, and 16. 



The typical form consists of a simple thread-like body slightly 

 clubbed at its free end. The filament shown in fig. 3 was nearly 

 •1 millimeter in length, and about "0012 millimeter in breadth 

 at the thickest place. The dimensions vary extremely ; for in- 

 stance, the thickness may be reduced to that of the very delicate 

 filament seen in fig. 6, whose length cannot be estimated because 

 of the complicated tangle in which it is arranged. Another and 

 far more elaborate knot of filaments measured '2 millimeter in 

 length ; if it was composed of a single filament, which is quite 

 possible, the latter would certainly have been 2 miUimeters in 

 length when uncoiled. Even the comparatively thick filament 

 in fig. 2 measures "4 millimeter in length. These entanglements 

 are of common occurrence, and usually consist of a labyrinth of 

 sweeping curves lying close up to the gland. Fig. 6 was 

 selected from the curious angular disposition of the coils. Again, 

 the thickness may be larger when compared with the length, as 

 in the smaller masses in figs. 2 and 3 ; and thick, partly dumb- 

 bell shaped masses may be combined so as to produce a branched 

 filament, as in fig. 16. In fig. 4 an almost spherical filament^ is 

 seen. In fig. 16, again, rather thick filaments are seen, one of 

 which forms a loop by having its distal end attached or closely 

 applied to the summit of the gland. The formation of loops is 

 an extremely common phenomenon ; they do not seem to be 



1 'Bot. Zeitung,'1868. 

 ' ' Annales des Sc. Nat.,' 1865. 



' The word filanieni is used to express any motile mass, of whatever 

 shape it may be, arising from a gland. 



