August 9. 1900] 



NATURE 



35: 



this kind are given by Worthington G. Smith i^loc. cit.), al- 

 though at the time (1884) these were raised, the imperfect 

 knowledge regarding the species of rust-fungi somewhat 

 weakens the arguments at the present time. Yet Eriksson, 

 Carleton, and other recent workers show even greater opposi- 

 tion. We are told that rust is a serious epidemic amongst 

 wheat in Australia where there are no native barberries ; in 

 India an undoubted black rust {Puccinia graminis) occurs on 

 wheat where there are no barberries nearer than 300 miles off 

 in the Himalayas. More conclusive is Eriksson's observation of 

 aecidium-cups on a barberry, yet the rust could only be traced 

 on rye, barley, and couch-grass to a distance from 10 to 25 

 metres ; or, again, tufts of Festtua elatior were found with ure- 

 dospores of Puccinia coronifera, while across a road there was 

 a hedge of Rhamnus almost free from the aecidial stage, yet 

 easily infected by means of teleutospores from the grass trans- 

 ferred to it by hand. These observations appear to show, 

 firstly, that the aecidial stage is not necessary in the life-history 

 of rust of cereals ; secondly, that the range of infection by 

 aecidiospores, or the reverse, is not great. What then remains 

 is one of the following possibilities: (l) The mycelium can 

 hibernate and resume activity in the following spring ; (2) the 

 fungus is carried over to a new season by the seed-grain, either 

 through adherent spores, or in some form internally ; (3) the 

 uredospores can hibernate ; (4) the teleutospores can infect other 

 Graviineae. In reviewing these we can only give a few of the 

 leading points relating to cereals. 



It is unlikely that the mycelium hibernates in the dead 

 remains of grasses, because it must then be capable of living as 

 a saprophyte for a short time in spring when it awakens to 

 activity ; this has not yet been proved. If, however, one 

 e.\amines the undergrowth of an area of grass in winter, 

 green shoots are generally present ; the mycelium may winter 

 here. Vet grain-crops are never sown except on ploughed land, 

 and there is no good evidence to show that epidemics of rust, 

 extending over whole acres, are propagated altogether from 

 patches of wild grass. Uredospores adhere to the grain of rusted 

 cereals, and there is a fair amount of evidence to show that 

 these may assist the fungus through the winter. In the 

 United States, Carleton believes that the uredospores of the 

 orange leaf rust of wheat and rye are produced and can ger- 

 minate late in the autumn, and so infect the sprouted autumn- 

 sown crop for next year. Eriksson, on the other hand, has 

 failed to discover that any one of the rusts of wheat lives all 

 round the year in the Uredo stage in Sweden, although Sorauer 

 states that in Germany the Uiedo mycelium of Puccinia riibigo- 

 vera hibernates without injury. There is thus a possibility that 

 the uredospore stage may transmit a rust from year to year. 

 This is more probable if the climate be suited to prolong the 

 growth of grasses late into autumn or early winter, and if the 

 specialised form of rust has more than one host- plant ; if it 

 occurs only on one cereal, it seems improbable that enough stray 

 plants are present after harvest to account for a widespread re- 

 appearance of rust in the succeeding year. Whether uredo- 

 spores adhering to straw or grain can survive the winter and 

 germinate has not yet been made quite clear. In laboratory 

 experiments it has been observed that uredospores frequently 

 exhibit deferred germination. After being soaked in water, only 

 a small percentage may produce germ-tubes. Eriksson and 

 others have observed that if the dormant spores are cooled in ice, 

 a further proportion are induced to germinate. Klebahn states 

 that the greater number of all forms of spore germinate if placed 

 on a suitable host-plant ; he believes that the proportion which 

 do not germinate at once, do so gradually later on, and sees in 

 this an adaptation for preservation of the race. 



The teleutospores ol the rusts of cereals have, as a rule, proved 

 incapable of infecting cereals or grases, the aecidium stage must 

 intervene. There is, however, the objection that observations 

 can only be made under more or less artificial conditions in 

 laboratory or green-house, and may not fulfil all the conditions 

 of infection out-of-doors. Plowright recorded an instance of 

 infection of cereals from the teleutospores of Puccinia graminis 

 (Gardeners' Chronicle, August 19, 1882), but gives no special 

 prominence to it in his '* British Uredinea; " (1889). Indirectly 

 certain facts lead one to suppose that teleutospores may have the 

 power to reproduce other stages in the life-history than the 

 aecidium. A rust-fungus of the group Leptopuccinia (e.g. Puc- 

 cinia ina/vacearu-n on mallow and hollyhock) produces only 

 teleutospores ; these give rise to sporidia, which re-infect the 

 mallow, and form the mycelium from which a new crop of teleuto- 



NO. 1606, VOL. 62] 



spores arises. Many forms of rust-fungi have only teleutospores 

 and uredospores on the same host : for instance, in Eriksson's 

 list many of the forms have no Aecidium ; either the hosts of 

 this stage remain to be discovered, or the teleutospore produc- 

 tion is fruitless, or the teleutospores are capable of bringing 

 about infection of the cereal or grass host. Species like Pucci- 

 nia suaveolens on thistles have all the forms of spore on one 

 host-plant except aecidiospores, which are unknown ; here the 

 teleutospores must, after hil^ernation, bring about re-infection of 

 the host. The existence of rust- fungi producing all the forms of 

 spore on one host, shows that two hosts are not necessary for the 

 development of the aecidium stage, and suggests that hetercecism 

 may be a later development in the history of the group. It is 

 also noteworthy that the teleutospore is produced by all rust- 

 fungi, with very few exceptions. It is therefore not improbable 

 that the teleutospores of hetercecious rust-fungi may still, through 

 their sporidia, retain the power of infecting the host on which 

 they are produced ; in other words, that heteroecism may be 

 facultative. 



In investigating the germinative power of teleutospores, 

 Eriksson finds the general opinion, that teleutospores must 

 hibernate, to be true only to a certain degree. As a rule, they do 

 not germinate unless they have passed the winter exposed to all 

 the changes of weather out-of-doors. Spores collected in autumn 

 and kept indoors soon lose the power of germination ; hence 

 his conclusion that rusted straw housed in barns will not be in a 

 condition to propagate the disease next spring. In the case of 

 spores left out-of-doors, the germinative power decreases rapidly 

 during the year after their formation, and in October they no 

 longer germinate. There is one exception, teleutospores of 

 Puccinia graminis tritici have a feeble power of germination 

 after two winters. On the other hand, Eriksson finds that cer- 

 tain teleutospores (e.g. Puccinia dispersa and P. ghimartim 

 tritici) can germinate in the year of their formation ; in the 

 case of the former species, aecidia were produced on Anchusa in 

 a short time ; in the latter form the host of the aecidia is un- 

 known. Plowright states that a bundle of rusted wheat straw 

 laid near ^\^ia\.% oi Anchusa in August 1885 produced aecidia 

 in September. 



Eriksson, after all his experiments, professes to be at a loss to 

 account for epidemics of rust on cereals year after year by ex- 

 ternal contagion alone, and he adopts the view that infection is 

 due to an internal germ. He thus introduces an agent for the 

 propagation of parasitic fungi which has hitherto been received 

 very sceptically by the plant-pathologist. His conclusion is the 

 result both of experiment and examination with the microscope. 

 In the experiments, varieties of cereals were used which are 

 known to be specially liable to rust. Vigorous shoots were taken 

 out-of-doors in spring, and enclosed in long glass tubes with 

 the open ends closed with cotton wool. Seeds were also 

 germinated in sterilised soil and kept in culture boxes, with pre- 

 cautions against entrance of spores by stuffing the ventilators 

 with cotton wool. At first the results were negative, but after 

 (in some way) improving the methods, the test-plants showed 

 rust, especially those shoots taken from out-of-doors. Examina- 

 tion with the microscope failed to reveal any mycelium or other 

 traces of the fungus until a few days before appearance of the 

 rust externally. At this time, however, with the aid of staining 

 reagents, certain protoplasmic bodies were observed in the green 

 cells near the margin of a rust- patch. These plaslids occur 

 solitary or in masses in a cell ; they are oblong or slightly 

 curved, simple or somewhat branched, and recall the form of 

 bacteroids found in root-tubercles of the Leguininosae. In a 

 short time the branch-processes pierce the cell-vvall of the host, 

 and develop outside the cell into an intercellular mycelium, 

 part of the original plastid' remaining inside the cell as the first 

 haustorium or sucker. Soon after this a rust-pustule appears on 

 the exterior of the host. The plastid is regarded as having 

 passed a period mingled with the cytoplasm of the host "in a 

 kind of symbiosis," till in response to external conditions— nu- 

 trition, moisture, heat and light — the " mycoplasm " becomes 

 separated from the cytoplasm, and assumes the form of a plastid. 

 The mycoplasm has its origin from the rust-fungus in the parent 

 host-plant, it becomes located in the embryo of the grain, de- 

 velops apace with the young plant, and so bridges the period 

 between one crop and the next; The following general facts 

 are said to support this mycoplasm theory : (i) The appearance 

 of rust on plants carefully isolated from contagion ; (2) the 

 disease in a field appears regularly four to five weeks after sow- 

 ing the grain of certain varieties of wheat and barley known to 



