14 THE LOIiANTHACEAE OP AUSTJRALIA^ 1., 



The mode of attachment of the adventitious roots depends largely upon th& 

 host. They occur more frequently, and are more vigorous, upon hosts with a 

 persistent bark. The swelling or ball-like growth is commonly associated with, 

 hosts having a deciduous bark. On one occasion I came across a plant of 

 Phrygilanthtis eucalypt if alius gi'owing on the trunk of Angophom lanceolata 

 Cav., which had no swelling where the union took place, but, instead, a deep 

 depression around the base of the parasite, as if it were gxowing from a hole 

 in the trunk of the host from which two strong adventitious roots descended for 

 about six feet; one was almost dead, and entirely free from the host, the other, 

 in a fairly healthy condition, had, at regular intervals of about six inches, sent 

 out haustoria from the lower surface, which had penetrated the host in the same 

 manner as a seedling plant. These had become elongated, and no less than 

 seven layers of the dead bark of the host were counted on two of them, and five 

 layers on three others. In two places I was able to jjass my fingers between, 

 tlie host and the main root. The largest rootlets were about an inch long and 

 conspicuously ringed. The old bark of the host had partly crumbled away, and 

 could be turned round and round like so many washers. The point of attach- 

 ment of the rootlets was much smaller and extremely Aveak. The struggle for 

 existence of this particular plant was noticeable in the sparseness of its foliage. 

 The annual shedding of the host bark was to all appearances detrimental to it, 

 as it loosened the feeders annually, thus depriving them of the necessary sup- 

 port. 



When adventitious roots a,re formed upon plants with a persistent bark, as 

 in the case of Angophora intermedia, Exocarx)us cupressiformis, Syncarpia 

 Itmrrfulia and many others, they are very firmly attached to the host. On old- 

 plants they form a matted mass encircling the host for some distance on either 

 side of the attachment, but usually they are more numerous on the side facing 

 the South. Occasional examples will be seen when the main root develops to a 

 considerable size of varying thickness. This applies more particularly to the 

 adventitious roots of P. celastroides when living upon Syncarpia laiirifolia.. 

 Sometimes the adventitious roots are free, and sway about in the wind, like 

 those formed on the branches of some species of Fiats, and continue to grow 

 until a. suitable object is reached for them to prey upon. It happens that they 

 sometimes reach the ground but do not penetrate it, as in the case with the roots 

 of the Ficiis, but become dry and withered at the end. 



I had under investigation at Turramurra, near Sydney, a plant of Phrygil- 

 anthus eucalyptifolius growing on the trunk of Acacia decurrens var. .mollis,. 

 with two adventitious roots touching the ground. I heaped some soil around 

 them to ascertain whether they would root in the soil, but in the course of two 

 months I found that the portion of the roots I had placed in the soil were dead,, 

 and above the dead part new roots had formed which were well beyond the- 

 level of the ground. This matter is worthy of further investigation in other 

 species. It, however, proves one point, — that the adventitious roots of Pliry- 

 gilanthus differ entirely from the so-called adventitious roots of other genera. 



The adventitious roots of this Family are capable of feeding upon each 

 other. It is a common occurrence to, see conglomerate masses of roots fused 

 together by their own haustoria, thus suggesting the idea, that the parasite 

 actually lives upon itself. When the roots are broken they produce new roots 

 from the side near the end of the broken part, and continue to grow as in the 

 first instance. 



Many of the roots are seen to produce young plants along tliom, but in. 



