46 



that why the seed spreads so well in South Australia., is the 

 long period of dry weather, accompanied by the strong hot 

 winds. This may free the seed from its woolly pappus, and 

 secure its transport spread and desiccation until it is ready 

 for germination. This may be exactly the time when the 

 autumn rains come in South Australia. None of these 

 conditions would be favourably found in New South Wales or 

 Queensland. December and January are wet months, indeed 

 all the summer is more or less rainy. The seed would 

 neither be scattered nor dried. The woolly covering would, 

 when wet, fix it on the soil. I only offer these suggestions, 

 not as giving the right explanation, but as sho^/ing how the 

 facts are capable of interpretation by a careful attention to 

 the habits of the plant. 



Another curious instance of the rapid but partial spread of 

 a weed is found in the same order of Comjjositce. This is the 

 well known Carcluus marianus, Linn. It is equally common 

 in the southern parts of South Australia, in Victoria, New 

 South Wales, Tasmania, and some parts of Queensland. It 

 chooses open ground for its spread, though very often poor and 

 sandy soils. It produces immense tufts of its peculiar large 

 juicy leaves. Sometimes it covers the ground closely — as 

 closely as it can possibly grow — and this to the extent of 

 several acres. It thus spreads into a thicket four or five feet 

 high, with the large flowerheads on their long stalks con- 

 siderably higher. In this way I have seen it keep to the 

 same open space for two or three years, not spreading much 

 beyond, and then I have known it as suddenly to disappear. 

 It seems to be a more troublesome and persistent weed in 

 Tasmania than in Australia. In New South Wales it is 

 regarded with a more friendly eye. In some portions of the 

 Batlmrst District, during the severe drought of 1876-77, 

 horses used to feed upon the seeds contained in the dry 

 flower-heads. In its green state both horses and cattle will 

 readily take to the young leaves and stumps when other feed 

 is scarce or dry. The way in which it suddenly appears and 

 spreads in remote portions of the bush is very singular. 

 Thus, on the Elderslie sheep station, just on the boundary 

 between the colonies of South Australia and Victoria, there 

 was an old sheep camp near Lake W^allace. I remember in 

 1857 noticing that this was covered with Cardicns marianus, 

 though there was none anywhere else in the neighbourhood. 

 It is a very local weed, and not troublesome, bat the singular 

 thing about it is the spreading so much in certain places and 

 no further. I have observed that it is very common all along 

 • the rich flats of the Lachlan River. 



Cardtms (cnicus) laoiceolatus, the so-called Scotch thistle, is so 



