130 ON AXOMOCL.VDA. 



butonly from the uader surfaceof the plant. What are called the rhiziaa, 

 or radicles, of llepatica), are capilliform, jointless tubes, proceeding in 

 fascicles or bands from the basal cells of the postical leaves (stipules or 

 araphigastria, as they have been called) ; or from the underside of the 

 stem itself, where there are no stipules ; but in plants like Anomoclada 

 relegated chiefly to the flagella. In terricoloue, corticolous, and folii- 

 colous species of prostrate habit, they are often present along the 

 whole length of the stem and branches ; but in some genera (e.g., 

 Plagiochila) they are almost confined to the prostrate caudex, and are 

 normally absent from the tall, branching, erect, or pendulous stems. 

 They vary in length from a small fraction of a millimetre in the 

 minuter species, to several inches in the gigantic Marchantia 

 paleacea of the volcanoes of the Andes. They exude mucus from their 

 extremity (which is often clubbed, knotty, or even slightly branclied), 

 and thereby adhere to the matrix, or to whatever crosses their path. 

 In the species of Radula, like R. Jlaccida, which overrun living leaves 

 in the tropics — close allies of the Jung, complanata, common on the 

 smooth bark of our trees — the rootlets spring from a conical depres- 

 sion on each semipostical leaf (or lobule) ; but very often the cones 

 themselves adhere to the matrix by their viscid apex, without the 

 intervention of any rootlets, and then they remind one of the propedes 

 of a caterpillar. Knowing all this, I naturally turned first to the 

 rootlets, where I found no abnormal mucosity ; but I saw plainly that 

 the mucus, which was poured out in such quantity as literally to 

 flood the entire plant, was derived directly from the stipules, whose 

 marginal and apical cells were continually swelling and discharging 

 their protoplasm, adlieriug for a while as empty bleached bladders 

 (jg millimetre in diameter), then falling away, fur the succeeding cells 

 to undergo the same process. It was not until after a long search that 

 I found a perfect stipule, which was broadly ovate with a subulate 

 point, and arose from a wide arcuate base, decurrent at each angle ; 

 but by this continuous disintegration the stipules are often reduced to 

 a 7Kirrow, semilunate rim, with a more or less ragged upper edge. 



The ])o>session of this viscid secretion renders the Anomoclada 

 eminently insecticidal, and it is the only Moss known to me which 

 has that property. The dead insects I used commonly to find in it 

 were mites, little flies, and (above all) minute spiders — which had 

 probably been entrapped as they alighted from some overhanging 

 branch, and on re-examining the specimens I find no others — not a 

 single ant— whence, although I did not taste the mucus, I conclude 

 it was not sweet, for, if it had been, some even of that sagacious tribe 

 would have been lured lo destruction. I saw no sign that the plant 

 snared the animals for food, any more than many other insect-trappers 

 do. On the llio !Negro the leaves of Droscra tenella, II.B.K., and 

 I), communis, St. Ilil., caught flics, rolled up over them, and doubtless 

 digested their soluble parts before again unrolling ; as has boon clearly 

 proved in the case of European species of the same genus. In a ravine 

 on the volcano Tunguragua I gathered Pinguicula cahjptrata, H.B.K., 

 and noted dead flies in the inrolled margins of the leaves, just as I 

 have often done in those of P. vulgaris in England ; but on bushes 

 overhanging the ravine grew a small Fern I^EIaphoglosaum glutinosum, 

 or an allied species), whose viscid frond-stalks slew far more insects 



