November 2, 1896.] 



KNOWLEDGE. 



243 



miniature volcanoes. At first the fecidium fruit is a small 

 spherical body formed beneath the epidermis of the leaf 

 ■whereon it grows, which it ultimately ruptures ; the 

 fecidium itself, when ripe, bursts, and the yellow spores are 

 discharged. The section of an iecidium shows a cup-like 

 cavity with the spores arranged in vertical rows like short 

 strings of beads ; they are developed by budding, and 

 become detached in succession. Externally the iecidium 

 is in most species invested by a membranous envelope, 

 the peridium, usually cup-shaped, but occasionally, as in 

 the cluster-cups of the pine, prolonged into a tube. The 

 peridium may open irregularly or split up in a definite 

 manner, giving its margin a toothed appearance. An 

 aecidiospore can germinate when sown on a suitable host. 

 The cluster-cups appear earlier in the season than the 

 uredo or telutospores, and are very often associated with 

 smaller cups called spermogonia, which appear on the 

 upper surface of the leaf (Fig. 5, 1 spm.), from which 

 issue minute spermatia, which have never been known to 

 germinate, and are therefore generally regarded as male 

 reproductive cells. 



All the three kinds of spores above described, it must 

 now be explained, are produced in succession by some of 

 the I'rediues on the same mycelium. The Puccinias of 

 the mint, primrose, violet, goat's-beard, and onion develop 

 all three forms ; teluto, uredo, and iecidiospores occur on 

 the same plant. Had we examined the bramble Phrag- 

 midium earlier in the season we should have found, not 

 the many-celled telutospores, but unicellular uredo or 



ipcidiospores. The rose 

 rust, PIi. subcortirum, and 

 that of the barren straw- 

 berry. Ph. frai/ari(s, in like 

 manner bear three kinds 

 of spore on the same host. 

 The rusts of the knot- 

 grass, beet, geranium, and 

 valerian, caused by species 

 of Uromyces, also possess 

 spores of three kinds. 

 Others, like U. alcheinilliB 

 and U. rumkis,hsiYe teluto 

 and uredo but no iecidio- 

 spores. Only telutospores 

 are known to be produced 

 by the Puccinias parasitic 

 on the gout-weed, speed- 

 well, miUow, harebell, and saxafrage. Uredospores are 

 wanting in the Puccinias of the ragwort and earth-nut ; 

 telutospores are absent in the rusts of the figwort and 

 fern, while neither the uredo nor telutospores are known 

 which correspond with the neidia of honeysuckle, meadow- 

 rue, and gooseberry. The three kinds of spore are not 

 formed simultaneously ; further observations may therefore 

 be expected to reduce the number of these exceptions. 

 Before it was known that a cluster-cup, a rust, and a brand 

 might be merely successive stages of the same fungus, 

 specific names had been assigned to eiioh of the forms, 

 with the result that soma of these parasites have three 

 names ; and this inconvenience is still unavoidable in cases 

 where the connection between the different stages has not 

 yet been demonstrated. 



But what invests this group of fungi with peculiar 

 interest is the fact that m;xny of them spend their iirst or 

 lecidium-bearing stage on a different species of host-plant 

 from that which they inhabit at a Liter period of their 

 life history, when they develop uredo and telutospores. 

 Thus there are several kinds which produce iieidia on the 

 leaves of firs and pines, and then migrate to plants of the 



Fio. 3. — Uredospores. 1. Portion of 

 grass blade, with rust. 2. Spoivs 

 of bramble rust. 3. Spores of 

 barren strawberry. 4 and 6. Spores 

 of corn rusts. 6. Spores of rose 

 rust. 7. Spores of thistle rust. 



Fio. 4.— _Ecidia. 1. leaf of berberry, 

 with cluster-cups. 2. Side view of 

 group of a;cidia. 3. Leaf of sunspurge 

 spotted with Melamapora euphoi-bi(f. 

 4. Cluster-cups of bedstraw seen with 

 lens. 



heath order. To this changing of hosts the name 

 Hetercecism(/i.^/, "other"; oM-os, "house") has been given. 

 Analogous phenomena are observed among animal para- 

 sites. The same organism which occasions "measles" 

 in pork, afterwards gives rise to the tapeworm in man ; 

 the tapeworm of the cat is but a more advanced form of one 

 that inhabits the intes- 

 tines of the mouse ; 

 and the liver fluke 

 of the sheep passes 

 one part of the cycle 

 of its development in 

 the body of a pond 

 snaU. Farmers long 

 suspected that the 

 presence of berberry 

 bushes in their hedges 

 had something to do 

 with the rust that de- 

 stroyed their wheat. 

 This idea was verified 

 by the discovery that 

 Pnccinid rjraminis is 

 merely a later stage 



in the development of AUcidium berberidis which infests 

 the berberry. As the alternation of generations was 

 first traced in this species, it is the example of 

 heteropcism usually given in text-books, but a similar 

 connection has been made out in many other instances. 

 The cluster-cups of the Scotch fir belong to the 

 same Uredine which bears teluto and uredospores on 

 the groundsel ; those of the colt's-foot correspond to 

 telutospores on the meadow grass of Pi(ccinia pminm ; 

 .T'.cidium uiiic(B of the nettle develops uredospores on 

 species of Carex ; the iecidium fruits of (TiiiininsjNinni'iiuiii 

 cdiirt'lliitii occur only on the leaves of the mountain ash 

 and other Pomaceie, the telutospores only upon those of 

 species of juniper. The ii'cidium of the buckthorn is 

 related in the same way to Pwriiiiii nirtinntii, not uncommon 

 on grasses. Again, the ivcidia of the orchid, onion, dock, 

 and dandelion appear in their uredo forms on various 

 grasses and sedges, while the parasites of certain Composites 

 seem to migrate to other plants of the same order. The 

 corn rust, P. ruhiiiri cr,;, turns out to be the second stage 

 of an iicidium that grows on the leaves of Anchusa and 

 other plants of the borage family. 



From these examples it will be seen that in fungi of this 

 description each generation of each species has its own 

 form of fructification and its own peculiar host-plant. 

 The brands of the mint aud bramble are not hetera>cious, 

 but produce all three sorts of spore on the same host, or 

 even on the same mycelium ; the I'redmes of the honey- 

 suckle, meadow-rue, and gooseberry, of which only the 

 ii>3idium forms are known, are likewise restricted to one 

 species of host. In this country ./■.'. iirossiilaria only 

 produces n'sidiospores ; telutospores are stated to have 

 been observed on the gooseberry itself on the Continent. 

 Should this be confirmed, it would appear that the fungus 

 in question is confined during its whole existence to the 

 same plant, and does not, therefore, possess the heter*cismal 

 character. 



In the life history of oao of these migratory fungi we 

 have then the following phases : — The earliest form 

 inhabits the leaves of a plant such as the berberry, where 

 it exhausts its energies and completes its career by the 

 production and discharge of the ii'cidiospores ; the latter 

 are incapable of germinating on the berberry, but on being 

 transferred to wheat, at once germinate aud form a 

 mycelium which develops the uredo and telutospores. The 



