June 18, 1897]. 



SCIENCE. 



94c 



growths that attract insects — commonly 

 ants— by oifermg to them some bait, such 

 as waterj' liquid or honey. In return the 

 ants commonly act as a body-guard or gar- 

 rison, and thus protect the entertaining 

 plant from being browsed down by insect 

 enemies. From the investigations of Belt, 

 Beccari, Delpino,* Forbes, Schumman, 

 Treub and others, we now count the number 

 of these as at least 3,000. They fall, how- 

 ever, under two distinct categories : (1) 

 Domatia which are merely extra-floral 

 nectaries, and which may be modified stip- 

 ules, or may be outgrowths over twigs and 

 leaf surfaces. (2) Domatia in which the en- 

 tertained insects puncture or excavate some 

 part of the plant. As an outcome, definite 

 holes, cavities or galleries arise, from which 

 plant liquids may escape, or in which the in- 

 sect-garrison may be appropriately housed. 

 From the researches of Molliardf and 

 Rathayl it appears that transition combina- 

 tions — and some of them very funnj' — can 

 be traced from Cecidia to both kinds of 

 Domatia, but in all we have acquired char- 

 acters of a remarkable kind. It always 

 appeared to me peculiar and somewhat in- 

 explicable, that neither of these should 

 be, so far as our knowledge went, trans- 

 missible by seeds. One may cut open 

 thousands of Cecidia in their season — chiefly 

 spring — and always the swelling is found 

 to be tenanted by the inciting cause — the 

 insect. The publication of Lundstrom's 

 paper § was a welcome one therefore, for, 

 though we could wish for wider verification, 

 his statements seem to be cautiously made. 

 Experimenting with Rhamnus alaternus that 

 forms cavernous domatia inhabited by 

 mites, he found that seeds infested by the 

 mites produced seedlings on which the 

 animals propagated rapidly, and at once 



*Mem. E. Accad. Sc. 1st. Bologna, Vol. 8, 1888. 

 tAnn. des So. Nat., Vol. 1, n. s. 1895. 

 tSitz. K. K. Zool. Bot. Gesell. Wien, 41, 1891. 

 § Nova acta K. S. Sc. Upsal, Vol. 13, 1886. 



formed their burrows, while other seeds, 

 selected and cleaned, gave rise to plants 

 that at first showed no signs of domatia, 

 and had no mites, but the later-formed 

 leaves developed the domatia as usual in the 

 axils of the leaf veins. These were smaller 

 and poorer in hairs than the normal 

 growths, but showed no trace of mite. 



Equally striking is the history of the 

 well-known Javan plant, Myrmecodia tuber- 

 osa, and even if we accept with Treub that 

 a small swelling and water canals exist in 

 the unpunctured swelling it seems to me 

 extremely likel}' that this is acquired and 

 hereditary. My reasons for this opinion 

 are founded on the history of a Bornean 

 pitcher plant. Nepenthes bicalcarata, which I 

 have studied from cultivated and from 

 dried imported leaves. First introduced 

 to science and cultivation by Mr. Burbidge, 

 it is now grown in collections and thrives 

 well. At the junction of the tendril with the 

 pitcher and parallel to the latter is an elon- 

 gated fusiform swelling. About the middle 

 it is pierced, in the wild state, by a neat circu- 

 lar orifice that leads into a cavity resulting 

 from breaking down of soft, loose, water-con- 

 ducting cells. In the cavity ants reside, and 

 can safely sip the juice that percolates from 

 the liquid-filled pitcher — cavity alongside. 

 I have examined a considerable number of 

 cultivated plants, and on every leaf was a 

 decided swelling filled internally with soft 

 cells. No other species of Nepenthes ex- 

 hibits such an enlargement. Whether culti- 

 vated seeds would reproduce the acquired 

 character we are not j'et in a position to say. 



As regards Cecidia, we know that these 

 arise, not immediately after the plant tissues 

 have been punctured by the insect when 

 ovipositing, nor after the egg has been de- 

 posited, but only when a larva hatches and 

 exudes some specific irritant. That this 

 irritant should, nevertheless, start in the 

 plant a formation of embryonic tissue that 

 develops in as definite a manner as if it 



