210 
correspondence of the foregoing results with those obtained by 
Vernon in echinoderms. 
The difference between different members of the same family 
must be in part attributed to the potential difference of the cells 
from which they are respectively developed. Whether the 
reducing division of the germinal cells is qualitative as well as 
quantitative is an open question, but there is reason to think that 
the life-history of these cells previous to conjugation may give 
opportunity for environmental variation like that of the Protozoa. 
Turning now to the subject of ‘‘ germinal variation,” the author 
points out that the existence of such environmental variation in 
the germ-cells, apart from reducing division, together with the 
physiological differences dependent on diverse conditions of vigour 
and maturity, may be expected in most cases to preclude the 
new individual from assuming an exactly intermediate position 
between its parents. When the male and female germ-cells 
unite, a series of contests takes place between groups of vital 
units, the issue being decided by their respective qualities, 
individuality or character. 
When different varieties or species are intercrossed, the effects 
may differ not only in degree but also in kind from those of 
ordinary cross-fertilisation. The following are some of the 
results that have been obtained experimentally from such inter- 
crossing :— 
(1) The offspring may be almost exactly intermediate between 
the parents. 
(2) The offspring may resemble one of the parents and not 
the other. Thisis often the case when wild animals are crossed 
with tame varieties of the same species. (It must, however, be 
remembered that the resemblance may be only superficial, as was 
clearly the case in the experiments with half-wild rabbits cited 
above. ) 
(3) Some of the offspring may resemble one parent, some the 
other. (This seems especially likely to occur if one or both of 
the parents is a sport. Standfuss’s results with insects are in 
accord with this.) 
(4) The offspring may combine, almost unimpaired, the more 
striking characters of both breeds. This has been seen in both 
pigeons and rabbits. 
(5) New or unexpected characters may appear in the progeny. 
Three out of four of a litter of cross-bred rabbits developed a 
habit of ‘* spinning.” 
(6) When half-breeds are crossed, the offspring tend to be 
extremely variable. Evidence of this is plentiful both in animals 
and plants. 
(7) Sometimes the offspring, instead of resembling the parents, 
resemble former ancestors. Prof. Ewart mated a cross between 
an “archangel” and an ‘‘owl” pigeon witha white fantail. The 
issue was a bird with a striking resemblance to the ancestral 
“‘blue-rock.” (Analogous results have been several times 
obtained in the case of insects.) 
Prof. Ewart’s paper is interesting and suggestive to a high 
degree. It would be hard to overestimate the value of the 
experiments which he is conducting with so much care and 
judgment in his well-selected menagerie at Penycuik. 
A. D. 
RUST-FUNGUS. 
ROF. MARSHALL WARD’S investigations into the rela- 
tions between host and parasite in the case of the Brome- 
grasses and their rust-fungus are bringing to light some 
interesting facts which have important bearings on the long- 
vexed questions of wheat-rust and the rust problem generally, 
which, as is now well known, have passed into an acute stage 
of late, principally owing to. Eriksson’s enunciation of his 
belief that the fungus can be transmitted in an invisible form 
vid the seed. 
In addition to testing this mycoplasm hypothesis of Eriksson’s, 
the researches undertaken by Prof. Marshall Ward are also 
directed to put to the proof the questions of degrees of 
specialised parasitism raised during the last decade by the 
researches of Plowright, Kleebahn, Eriksson, Magnus, Fischer, 
and others, and more especially, to see if any deeper insight can 
be obtained into the causes of epidemics and the relative pre- 
disposition or immunity of certain plants to attack. 
In a paper read to the Cambridge Philosophical Society on 
January 20, 1902, Prof. Ward gave a summary of his results 
with more than eighteen hundred infection experiments, made 
NO. 1704, VOL. 66] 
NATURE 
[JUNE 26, 1902 
on twenty-two species and varieties of Bromus with the 
Uredospores of Puccinta dispersa (Erikss.), the brown-rust of 
these grasses. These results show clearly that, other conditions 
being the same, the infection of a given species of Bromus—say 
B. mollis—by the Uredospores of the Puccinia depends on the 
origin of the spores, that is to say, on the circumstances of 
nutrition and breeding generally to which they have been 
hitherto accustomed. For instance, if the spores are reared on 
B. mollis, they infect another plant of &. mollis readily ; but 
if they are reared on B, sterz/is, they refuse to infect B. mollis, 
though they will readily infect another plant of 2. sterd/is, 
But, in addition to the infective capacity of the spores con- 
ditioned by their past history, there is the question of the pre- 
disposition or immunity of the host. For instance, it is easy to 
infect Bromus mollis with spores from &. mol/is, but far less 
easy to infect B. vacemosus with such spores, and practically 
impossible to successfully infect B, steri/zs. Part of Prof. 
Marshall Ward’s work goes to prove that the immunity of 
given species of Bromus is not due to anatomical peculiarities, 
such as the number and size of the stomata, hairs, the volume 
of chlorophyll tissue and so forth, but to some substances or 
conditions in the living cells which escape microscopic investi- 
gation. In other words, the inquiry is being pushed into the 
domain of enzyme reactions, anti-toxins and so forth. 
In a forthcoming paper it will also be shown that the external 
conditions of germination of the spores, and of infection by way 
of the stomata, require far more attention than they have yet 
received. 
Ina paper read to the Royal Society on February 20 last, 
another aspect of the investigation was opened up, namely, the 
possibility of obtaining pure cultures of these Uredines, a method 
which applies to other parasitic fungi as well. . . 
In order to obtain more decisive answers to such questions as 
—Are any of the results obtained on plants in the open, or merely 
covered with bell-jars and so forth, due to spores accidentally 
introduced, or to mycelium, &c., already in the plant? a 
number of infections were made on seedlings germinated and 
grown antiseptically in tubes as follows :— 
Clean picked seeds were placed singly, by means of forceps, 
on filter paper at the bottom of Petri-dishes properly sterilised 
by heat. When these had germinated, and observation showed 
that the whole series was free of moulds or other signs of con- 
tamination, the seedlings were removed by means of sterile © 
forceps and transplanted singly into sterilised tubes of various 
kinds as described below, and the further growth allowed to 
proceed in the light under conditions varied as will be seen. 
Prof. Ward had already shown that seedlings will continue to 
grow in such tubes, but, as we have seen, in the cases previously 
described he had no guarantee that the seedlings introduced into 
the culture-tubes did not already carry on their leaves wind- 
borne or otherwise transmitted spores. 
In the case of these seedlings germinated from clean ‘‘ seed” 
in sterile dishes and tubes, it is obvious that the only chance of 
infection depends on spores attached to the ‘‘seed” or on 
mycelium in the seed. 
Experiments with seed gathered even from badly rusted plants 
and germinated as above have never given rusted seedlings, 
although other experiments have shown that the germ-tubes 
of attached spores can infect seedlings when the plumule is only 
3-5 mm. high. Nor has Prof. Ward ever been able to discover 
any trace of mycelium in the seeds. 
But if the ‘‘ seed” of the Bromus is sterilised before germina- 
tion—as can be done by steeping in various antiseptics, or by 
heating to 60-70° C.—it is found that pure cultures of the Brome 
may be obtained in the tubes, and it is then only necessary to 
infect such a clean seedling with the spores of the parasite to 
obtain a pure culture of the latter. 
Preliminary experiments soon showed that the Brome seed- 
lings thus raised from seeds treated antiseptically, and protected 
from the first by glass, may be grown for weeks and even for a 
couple of months in such tubes under proper precautions, and 
Prof. Ward set himself the task of ascertaining how ‘such 
cultures would behave in infection experiments. 
In the following experiment upright tubes of the kind known 
to chemists as ‘‘drying towers” were prepared as in the 
diagram (Fig. 1), so that by means of an aspirator attached to 
the tubing at G, a continuous current of damp air could be 
slowly drawn through the whole series, aérating the roots of 
the seedlings F, which burrowed into the cotton-wool 8B, day 
and night. The tubes were charged each with one seedling, 
