(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Orchid mycorrhiza"

ALBERT R. MANN 
LIBRARY 

AT 

CORNELL UNIVERSITY 




a0s^ 



jf ty"T ^ A,-^.-^ — jSrr: 



'^ 



Orchid Mycorrhiza 



J. Ramsbottom 



Reprinted from 

Charlesworth & Co.s Catalogue 
1922 



From the 

KNUDSON MEMORIAL 

COLLEGllON 

Botany Department 
Cornell University 



1^2 




Cornell University 
Library 



The original of this book is in 
the Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 



http://www.archive.org/details/cu31924000641492 



Orchid Mycorrhiza 

'By 



6^ 

J. Ramsbottom /^ 



Reprinted from 

Charlesworth & Co.'s Catalogue 
1922 



377830 



ORCHID MYCORRHIZA 

By J. Rams bottom. 



Introduction. 

One of the most interesting phenomena in biology is that generally known as symbiosis — the living 
together of two organisms in close association. It is usually considered that this intimate relationship is 
of benefit to both components. Many examples occur in the plant kingdom. The lichen is probably the 
best known of these, being a composite plant formed of a fungus and an alga in definite union. Other 
well-known examples are the bacteria (Pseudomonas radicicold) living in the root nodules of Leguminosae, 
and the Ginger-beer plant,' of which the lumps are composed of a yeast (Saccharomyces pyriformis) and a 
bacterium {Bacterium vermiforme). An intimate union can also occur between plant and animal, as in the 
case of the marine worm Convoluta, in the body of which an alga is always present, and as in the larv« of 
certain aphids, coccids, etc., where yeasts occur as more or less definite structures. 

Mycorrhiza. 

That the roots of many plants have the threads or mycelia of fungi associated with them has become 
very well known during the last eighty years. It is of interest to find that cells containing fungi were first 
figured in an orchid (though not very clearly) by Link^ in 1840, who observed them in the young seedling 
(protocorm) of Goodyera frocera. He did not hazard a guess as to their nature — his idea being that the cells 
were filled with colourless granular material which finally disappeared. 

At the beginning of the forties of last century the naturalists of this country who were curious in botany 
were very interested as to whether Monotropa Hypopitys was parasitic on the roots of beech in a manner 
similar to Lathraa. In 1842 we have T. G. Rylands' writing " On the nature of the byssoid substance 
found investing the roots of Monotropa Hypopitys." Rylands concludes that " the ' byssoid substance ' 
is really fungoid, and performs no essential function in the economy of the Monotropa." It is, however, to 
Reissek* (1847) that we owe our first real knowledge. He examined numerous plants and came to the 
conclusion that fungi were normally present within the cortical cells of the roots of various flowering plants, 
being best developed in the underground roots of orchids. In these he studied most of the native and 
several exotic genera. He found that in Orchis Morio, for example, the fungus was present in almost all 
the cortical cells, whereas in the tropical species the fungal masses were arranged singly at the periphery. 
The presence of fungi was most frequent in underground roots, less usual in superficial ones and very rare 
in aerial roots exposed to the light. Moreover Reissek attempted to extract the fungus from the roots. In 
those days of imperfect technique it is not so surprising that he failed as that he should have made the 
attempt. The fungus he obtained he named Fusisporium endorhizum : it is probably one of the common 
saprophytic species of Fusarium so abundant in soils. 

Another type of association between fungus and root is also well-known, particularly in forest trees. 
Here the fungus mycelium forms a sort of mantle round the root, in contrast to being within the cells of 
the cortex. Apparently Hartig first noted this type in 1 840 in the extremities of the rootlets of Pinus 
sylvestris although he mistook the hyphse for branched intercellular canals surrounding the internal cells 
such as are known to exist in the corky layer of the root cortex in Juniperus and Thuja. Rootlets so 
infected are most frequently coralloid in appearance. Gasparini in 1856 noted that such rootlets in 
Castanea and Corylus were surrounded by fungal hyphse. 

The term mycorrhiza was coined by Frank" in 1885 for the fungus-roots. Even at that date it was 
known that in some plants the fungus occurred in rhizomes as well as roots {f.g., Neottia), and since then 

'The Ginger-beer plant is, at the present time, being widely distributed over the country as " Califomian Bees," " Mace- 
donian (Salonika) Bees," " Mesopotamian Bees," " Palestine (Jerusalem) Bees," " Wine Bees," " Water Bees," " Balm of Gilead," 
etc. 

^ H. F. Link. Icones selectae anatomico-botanicae. II. p. lo. t. VII. (1840.) 

^ T. G. Rylands. On the nature of the byssoid substance found investing the roots of Monotropa Hypopitys. Phytologist. 
I. pp. 34J-8 (1842.) 

* S. Reissek. Uber Endophyten der Pflanzenzelle, eine gesetzmassige den Samenfaden oder beweglichen Spiralfasem analoge 
Erscheinung. Naturwiss. Abhandl. von W. Haidinger. I. pp. 31-46 (1847). 

* A. B. Frank. Ueber die auf Wurzelsymbiose beruhende Ernahrung gewisser Baume durch unterirdische Pilze. Ber. d. 
deutsch. bot. Gesell. III. pp. 128-145. (1885). Lehrbuch der Botanik. Bd. I. (1892), p. 264. 



many cases have been found for which the term is quite a misnomer {e.g.. Liverworts). It is a convenient 
term, however, and it is better to accept it with an extended meaning rather than to restrict it to those 
cases for which it is etymologically sound. Frank gave special names to the two types mentioned above. 
He used the term endoirophic mycorrhiza for those forms in which the fungus occurred within the tissues 
of the host and the term ectotrophic mycorrhiza where the fungus hyphx surrounded the rootlet as a sheath. 
These are convenient general terms, but it is well to remember that the two types are not absolutely distinct, 
as is seen, for example, in Monotropa, which had been well-described by Kamienski in 1883. Mycorrhizas, 
mainly endotrophic, have been described, either as usual, or occasional, in various Liverworts, Mosses, 
Horsetails, Club Mosses, Adder's Tongues, Ferns, Conifers and Flowering Plants : and in Algae apart from 
Lichens we have cases of constant association of fungi and seaweeds, as, for example, in Ascophyllum and 
Pelvetia, which each have their attendant Mycosphcerdla. The antiquity of such associations is seen in the 
fact that they occur in the fossil plants Rhynia, Hornea and Asteroxylon from the Lower (or Middle) Devonian 
— vascular cryptogams which from their simple structure and age are of the greatest theoretical importance. 
Weiss (1904) moreover recorded mycorrhiza in fossil roots from the lower Coal measures for which he pro- 
posed the name Mycorrhizonium, and Osborn (1909) found fungus mycelia in the inner cortex of Amyelon 
radicans the root of Cordaites. 

Orchid Roots. 

As we have seen, fungi have been recognised in the roots of orchids since 1847. A transverse section 
of an infected root taken just above the root-cap shows the fungus in the cortical cells. (Fig. I.) The 
distribution is more or less constant in the same orchid, but varies in different genera. It is only in the 
young root where root-hairs are present that the fungus is, as a rule, recognisable as such. The epider- 
mal cells are not infected. The fungus usually enters the root through the root-hairs, but in some species 
it apparently is able to make use of any portion of the piliferous layer. The hyphse' pass through the 
external layers to a more or less definite zone, where they reach their maximum development, rapidly 
spreading and completely filling the cells. If an exodermis be present the hyphae pass through the thin- 
walled transfusion or passage cells. The first two or three cortical layers of the root are thus generally free 
from fungus except where the hyphas of infection pass through them : even in these there is no balling of 
the mycelium in the cells. In some genera (Habenaria) (Figs. I and 14), the fungal zone occupies roughly 
the third and fourth layer of cortical cells. In other genera {Neottia and Epipogum) the fungal zone of the 
root occupies three layers or so of cells separated from the endodermis by about half-a-dozen cell rows. 
In other cases practically the whole of the cortex is occupied (Cymbidium and Odontoglossum). The central 
stele is never infected, the mycelium not entering the endodermis. The fungus also never infects the cells 
of the growing point of the root. Infected roots do not always show the endophyte in all their length, 
neither is it invariably present in a continuous zone. Infection does not generally occur once for all, but the 
hyphae from the soil infect the roots in several places and if the fungal zone be bf several cells thickness it 
is frequently seen as patches in transverse section. Nor, as a rule, are all the roots of an orchid infected. 
Aerial roots particularly are free from fungus, the only exceptions being where the roots are applied to the 
soil and are without chlorophyll. Such a case is shown in Fig. 15. Aerial roots can sometimes be found 
in such a position with the exposed portion green : in these circumstances if infection occur the fungus is 
restricted in distribution to the portion of the root without chlorophyll. In addition to cells containing 
chlorophyll those containing tannin, mucus, raphides and other crystals are never invaded by the fungus. 
Lateral roots are more frequently infected than main roots and in those genera with numerous roots (Orchis, 
Ophrys) according to Stahl only one out of three of the roots arising from the rhizome have fungus present in 
their cortex. Moreover, certain genera such as Listera and Epipactis, which have their chlorophyll particu- 
larly developed, seem to be irregularly infected, whereas plants poor in chlorophyll, e.g., Limodorum and 
Corallorhiza are well fungussed. All orchids so far investigated possess mycorrhiza, with the single exception 
of the saprophytic Wulhchlagelia aphylla.'' Large numbers both of native and exotic species have been 
studied — ^Wahrlich,' for example, examined over 500 of the latter cultivated at Moscow. 

Since the earliest workers, e.g., Reissek, it has been known that in some cells at least the fungus 
becomes changed from its original thread-like structure into glary yellow amorphous masses. In fact, it was 
owing to this phenomenon that the fungal nature of the cell infections of these roots was not at first generally 



• Janse has shown that in Lecanorchis jmanica the infecting hyphse are sometimes united into a mycelial ribbon. 

' Further investigation is needed on this plant. MacDougal first recorded that Cephalanthera oregana was free fron 
; later found a somewhat sparse and intermittent infection. 

• W. K. Wahrlich. Beitrage zur Kenntnis der Orchideenwurzelpilze. Bot. Zeit. XLIV. pp. 481, 497 (1S86). 



Ul. 



realized. Wahrlich paid special attention to the changes which took place and most investigators of orchid 
roots since then have taken note of them. Magnus' working with Neottia in which the alterations are well 
marked gave a clear description of the metamorphosis. He distinguished two main types of infected cells 
and held that there were no transitional stages. In the one type which he calls " digesting cells' 
(Verdauun^szellen) the fungus always degenerates ; in the other type, the " host cells " (Pilzwirthszellen) 
the fungus remains alive in the cells which lodge it and is thus able to hibernate. Magnus states that 
Neottia shows a more or less definite arrangement of these two types of cells, the digesting cells forming an 
outer and an inner and the host cells the middle layer. Such a regular arrangement is not usual in orchids — , ^ 
even in Neottia it is doubtful — and host cells are absent in certain native genera,such as Goodyera, and in most 
tropical forms. Bernard and Burgeff have also studied the question of the fungus digestion — the former 
mainly in seedlings, the latter principally in the root of Platanthera chlorantha. Before a hypha enters a 
host cell the nucleus of the latter increases in size. This action at a distance is also seen in the fact that 
starch disappears from the cells. The nucleus in the neighbourhood of the hypha becomes hypertrophied, 
often becomes modified in form and has increased attraction for stains. Where mycelial influence is great 
the nucleus becomes amoeboid and sometimes disintegrates : this would seem to indicate a parasitic action 
on the part of the fungus. The digesting cells are clearly recognisable by the degenerating mass which more 
than half fiUs the cells. The increase in the size of the nucleus is also a character as it becomes about four 
times its original diameter, i.e., roughly sixty times the volume. The nuclei become amoeboid and put out 
pseudopodia which serve to attack the hypha*;* The hyphae only stain slightly : they increase in diameter 
up to about double and also in length. ' The development in some cases is so great that the cell is quite filled 
with the thick mycelial mass and the nucleus is crumpled by the hyphse. Enclosed by the pseudopodia 
the latter gradually lose their outline until frequently they cannot be distinguished from protoplasmic 
trabeculas. The victorious nucleus then assumes a round form and normal volume and reconstitutes its 
chromatin network. The endophyte is reduced to an amorphous yellowish clump with indistinct contour, 
and is absolutely devoid of life : it is surrounded by a cellulose membrane. It would seem that as the root 
ages the clumps finally disappear. After the formation of the clump starch often reappears in the cell. 
Burgeff states that the fungus in the host cells can re-attack digesting cells when similar stages are again 
gone through. We shall return to the question of digestion when we consider the seedling. 

Germination of Seeds. 

The difiiculty in germinating the seeds of orchids is one which has been known for a considerable number 
of years. In fact, it was not until 1804 that any orchid seedhngs were described, when R. A. Salisbury 
figured those of Orchis Morio and Limodorum verecundum. Later, many botanists such as Link, Irmisch, 
Beer, etc., added to our information concerning the stages of development. Orchid growers evolved the 
method of sowing seeds on the soil containing the parent plant,^" and it was in this manner, or some modifi- 
cation of it, that most of the hybrids known in horticulture were raised. The facts known, i.e., the difficulty 
in germinating seeds unless placed on orchid soil and the presence of fungi in the roots, led many to suspect 
that the fungus was concerned in some way with the success or failure of germination. 

We have mentioned that when Reissek recognised the fungal nature of the cell inclusions, he attempted 
to isolate them. This attempt, long before the days of bacteriological technique, was bound to end in 
failure and the fungus he isolated was a species of Fusarium, a genus which has been time and again pro- 
claimed as the consort of the orchid root. (The genus Nectria also has often been assumed to be the 
endophyte). It is to Noel Bernard that we are indebted for our chief knowledge of the facts of orchid 
germination. This brilliant young French investigator began his studies on mycorrhiza in 1899, and they 
extended until his death in 191 1. His first investigation was on the germination of Neottia, In 1902 in his 
thesis i,tude sur la tuberisation he mentions that orchid seeds can germinate only in the presence of the 
root fungus and that the seedling is infected from its earUest stages. Realizing the importance of this fact 
he turned his attention to investigating it thoroughly and following the various ramifications of the subject. 
Bernard's great work Uivolution dans la symbiose. Les OrchidSes et leur Champignons commensaux, 
appeared in 1909. In the same year a comprehensive work by Burgeff was published entitled Pie 
WuTzelpilze der Orchideen. Both these investigators succeeded in isolating the fungus from the orchid root 
and growing it on nutrient media. Orchid seeds germinated without difficulty on having the appropriate 
fungus supplied to them. In describing the course of events full use has been made of the work of Bernard 
and Burgefi this being supplemented by observations made by the late Mr. J. Charlesworth and the writer. 

' W. Magnus. Studien an der endotrophen Mycorrhiza von Neottia Nidus-avis L. Jahr. f. wissensch. Bot. XXXV. 
pp. 205-272 (1900). 

"•I believe Dotniny of Messrs. Veitch & Sons introduced this practice. 

iv. 



Joseph Charlesworth (1851-1920). 

It will probably not be considered out of place here if I venture upon a few remarks concerning my friend 
the late ^T. Joseph Charlesworth. In the year 1913 I was invited to Haywards Heath to see his results in 
raising seedlings by what he styled the " pure culture method." He had succeeded in eliminating many 
sources of error and had achieved remarkable and consistent results in raising Odontoglossum and its allies 
by sowing seeds on nutrient media in which the appropriate fungus was growing. The probability that the 
mycorrhizal fungus in some way affected the germination of orchid seeds had influenced him for many years 
and he had earned his great reputation as a hybridist by his success in raising hybrids by modifications of 
the methods in common use. In an account of a visit to his establishment in 1906 it was written " Here 
is a veritable seedling land, thousands and thousands of them," and in 1909 " The raising of Odontoglossum 
and allied genera has become a very important business, and there are thousands of seedlings in existence. 
Messrs. Charlesworth are reducing it to a system." It was to one so successful by the older methods that 
Bernard's work made such a strong appeal, and he eventually decided to adopt the system. His culture 
flasks were sufficient testimony to the success of the laboratory method when placed upon a commercial 
scale. One was not prepared to find, however, that at the same time he had, after the age of sixty, become 
so embued with the new spirit as to have purchased microscopes, microtomes, ovens, stains, books, etc., and 
become proficient in microscopic technique. (The photomicrographs at the end of this paper are all taken 
from his preparations.) Naturally he did not restrict his newly acquired activities to studying orchids, 
but the main part of his laboratory work dealt with them, and he was especially interested in the seed from 
its first formation and in the relations between fungus and seed in germination. The whole of the slides 
were generously placed at my disposal. We, however, drew up a scheme of collaboration and mapped out 
a series of investigations, which unfortunately had to be discontinued owing directly and indirectly to the war. 
When, last year, we were both again free to resume the work he was a sick man and beyond application 
to research. 

I should wish to repeat here for the benefit of those orchid lovers who knew so well one part of his 
accomplishments that the other part was equally good. The fact that he should commence laboratory work 
at such a late age is as surprising as is the success which he attained practically unaided. To a botanist 
trained in the schools many of his expressions appeared whimsical, but when he termed the small cells at 
the distal end of an Odontoglossum seed the " soul of the plant," it was as a result of finding that it was 
there eventually that both stem and root were laid down — and he had a happy knack of coining such 
expressions and, one may add, a certain persistency in using them. If his early days had been spent in 
acquiring a knowledge of academic botany rather than in connection with his father's wool business, there 
can be no doubt that the name of Joseph Charlesworth would have been writ large in the annals of British 
science. As the firm of Messrs. Charlesworth are carrying on the traditions of their late chief it may be 
possible at some future date to complete and put on record, certain of the investigations ; and it is hoped 
it may be possible to carry out the original plan in which his knowledge of orchid culture would have played 
an essential part. 

Orchid Fungus. 

Bernard in his first attempts to isolate the fungus from orchid roots obtained a species of Fusarium. 
When, however, he succeeded in extracting the right fungus he established a criterion which enables one 
to settle without doubt whether the true fungus has been isolated, viz., that the endophyte is able to bring 
about the germination of the seed. 

The fungus, when living within the cells of the plant, shows no characters which give a clue to its sys- 
tematic position, but when it is grown on nutrient media it shows additional stages of development which 
are characteristic. 

When extracted from the root and placed in a culture medium the fungus always appears to behave 
in the same way. The fungus spreads over the surface by the apical growth of its septate filaments. Mean- 
while lateral branches arise and anastomoses take place between the hyphae. Later, balls of hyphse appear 
here and there in the culture and on the sides of the tube or flask containing them, usually some distance 
from the ends of the hyphse. These balls are very similar to those which appear in the cells of the root, being 
formed by the rolling up of the ends of young growing filaments, and often becoming very compact. When 
seen in the host cells this method of growth suggests adaptation to the needs of the special environment and 
its presence in cultures might lead to the supposition that the character is so impressed upon the fungus that 
it also shows it when living free. The character is, however, not rare in the group in which we must classify 
this fungus. 

As the mycelium becomes older shorter filaments arise with very short and swollen segments, which are 



apparently rich in food reserve. (It was this appearance that caused Bernard to place the fungus in the 
genus Oospora when he first studied it). These filaments ramify abundantly and in certain forms anastomose 
amongst themselves and give rise to yellow or brown sclerotia^' (Figs. 4 and 5) small spherical bodies formed 
of intertwined and massed hyphae. These structures are capable of withstanding drought and other 
inclement conditions and are remarkably tenacious of life. Bernard has pointed out that these swollen 
filaments are very like those which occur in Rhizoctonia violacea Tul.'^ which is common on potatoes, lucerne 
and other crops, where it forms small, blackish, irregular sclerotia, and he considers that the orchid fungi fall 
into the same genus. He classed the fungi obtained from about twenty orchids as three species, Rhizoctonia 
repens, R. mucoroides and R. lanuginosa. The first, which was by far the most commonly isolated {Lalia, 
Lcelio-Cattleya, Spiranthes, Paphiopeiilum, Cymbidium, Aerides, Bletiella, Ccelogyne), does not form 
sclerotia. R. mucoroides was found in Phalienopsis and Vanda, and R. lanuginosa in Odontoglossum (Figs. 
4 and 5). Burgeff, unaware of Bernard's latest results, proposed a new genus Orcheomyces for the reception 
of the orchid fungi. He fully describes fifteen species, naming them after the orchid from which he obtained 
them and mentions another fourteen by name : he divides them into five main groups. 

A discussion of the different systematic interpretations given by Bernard and Burgefl would be out of 
place here and for convenience the more generally adopted name Rhizoctonia will be used. The diversity 
in the number of species is simply a case of the usual " lumping " and " splitting." Bernard found in his 
experiments that fungi obtained from different sources, but to which he gave the same specific names, varied 
somewhat in their behaviour, and it is quite probable that these physiological distinctions are related to 
slight morphological differences. Bernard later recognised certain of Burgeff's species as falling within his, 
e.g., Orcheomyces Sambucina, 0. mascula, 0. insignis and (X.<Luddigi were regarded by him as Rhizoctonia 
repens — but he apparently took into account merely the gross characters of growth. 

The endophytic fungus is able to ferment cellulose, which accounts for its ability to penetrate cell walls. 
Burgeff made a study of the physiological charcters of the species he isolated. He found that they were 
able to absorb carbohydrates in the form of sugars, these being in all cases transformed by a diastase-invertase 
in some species, maltase in others. Having regard to the prevalent ideas as to the function of mycorrhizal 
fungi it is of particular interest to note that these forms are apparently unable to fix free nitrogen : the 
nitrogen of organic compounds, such as peptone, can be made use of as a source of nitrogen : ammonium 
compounds are better assimilated than nitrates. By growing cultures in the dark and in an atmosphere 
devoid of carbon-dioxide he established the fact that the carbon compounds of the soil can suffice as a source 
of carbon. , '" • ^. "- 

Bernard in his experiments found that the fungi if grown in culture gradually became inactive. Cultures 
two years old were quite unable to bring about germination. Burgeff, on the other hand, found that his 
cultures after twenty-six and twenty-eight months retained their power. In connection with this point a 
culture of a root fungus which had been regularly cultivated for at least eight years, though not used during 
that time for germination, was recently tried. A very feeble germination occurred in certain of the tubes. 
As the activity of the fungus when it was first isolated is not known, it is impossible to say whether there 
is any decrease in intensity, though this is probable. The gradual attenuation with final loss of activity 
noted by Bernard nxay be a consequence of " staling " through too infrequent renewal of cultures. He 
found that the intensity of an attenuated form can be increased by extracting it from a plant which it had 
been successful in germinating. 

Germination of Seeds— co«^;««erf. 

The seeds of orchids are very small, the embryo being frequently only just visible to the naked eye." 
They possess a single integument which is in the form of a characteristic network (Fig. 2), which varies 
somewhat in shape and structure in the different genera. On sectioning the seed (Fig. 3), or on viewing 
when stained and mounted whole there is seen to be no differentiation into cotyledon, stem, radicle, as is 
almost universal in flowering plants.** It appears to be most usual for the cells at the suspensor end of the 
seed to be somewhat larger than at the upper end (Fig. 3), though this is not always the case {Cypripedium). 
Sometimes the suspensor cells are permanent (Cattleya) — the suspensor is the stalk by which the developing 
seed is attached and nourished — at other times they disappear before the seed is matured (Phalienopsis'). 
Seeds taken from the capsule under sterile conditions and sown on ordinary substrata where no fungus is 

11 Sclerotia are known in all groups of fungi, often reaching considerable dimensions, e.g. the size of a man's head in Polyporui 
Mylittae (the " black fellows' bread " of Australia). 

*^ Corticium vagum B. & Br. yar. Solani Burt. 

'' The embryo in Fig. 2 is approximately zoofi i.e. c. jjj inches. 

'♦ Bletilla hyacinthina shows a rudimentary cotyledon according to Bernard. Treub has indicated a cotyledon in Sobralia 
macrantha, and Pfjtzer records a green embryo with a differentiated cotyledon in Platyclinis glumacea. 

vi. 



present do not as a rule develop. Generally they merely swell and become green (Odontoglossum) though 
sometimes even this does not happen {Epidenirurri) ; in other cases they may form stomata and the rudi- 
ments of hairs (Cattkya). The only case so far known in which any considerable development can take 
place under these conditions is Bletilla hyacinthina where Bernard found that thin slender seedlings developed 
with distinct leaves. The food reserve of orchid seeds is most frequently oil, part of which becomes trans- 
formed into starch. The reserve food comes to its end just as the seed commences to become green. This 
is usually after three or four months, during which time very little, if any, nutriment can be obtained from 
the substratum, as absorbing hairs are lacking. If no fungus infection take place then, the seedling dies. It 
is somewhat surprising that after the production of chlorophyll death should occur rather than autonomous 
growth by aid of photosynthesis : the seedling appears to form chlorophyll as a sort of last despairing effort. 

If, however, the appropriate fungus {i.e., the fungus from the root of the parent or some closely allied 
plant) be added now at the latest, an extraordinary change takes place. The fungus seems to give an 
impetus to development. 

In the culture flasks it is only in prearranged experiment that infection takes place at such a late stage. 
The fungus enters the seed usually within a few days. The course of events may be made out from the 
photomicrographs, which are taken from different genera in order to show the general similarity in the 
phenomena. Entry takes place at the suspensor end of the seed by the suspensor cells themselves, if such 
be retained. The cell walls here are unmodified, though the general surface of the seed is slightly cuticularized. 
As we have seen, the cells at the suspensor end of the seed are generally larger, and it is into these that the 
fungus passes. (Figs. 3, 6, 7, 8). The cells are invaded by degrees, the hyphae becoming twisted into a ball 
in each cell before passing on to the next. Almost immediately the smaller cells at the opposite end of the 
seed undergo division. It is here that the meristem of the stem is laid down. The meristematic cells in 
orchids are never entered by the fungus : the only cells capable of division which ever harbour the endophyte 
appear to be those of the seed where it first enters. Eventually the developing seedling takes on a swollen 
shape most frequently more or less turbinate (Figs. 9, 10). 

Bernard uses the term"protocorm" for this swollen tubercle and regards it as of theoretical importance, 
as it simulates the protocorms of Lycopods and the colourless underground prothaUi of Adder's Tongues, etc. 
It is of interest to remark that a similar structure, also associated with fungi, occurs in the primitive fossil 
plant Hornea from the Devonian. The fungus remains restricted to the larger cells and follows in the wake 
of their division. The epidermal layer is free from infection. Meanwhile the rapid division taking place 
in the smaller cells at the anterior end of the seed gives rise to the young stem apex and the first leaf 
(cotyledon). About the time this young leaf becomes visible to the naked eye the cell-division has become 
extended along the axis and the beginning of the central stele is seen (Fig. 10). In this manner the young 
root is formed and begins to absorb its way through the tissues of the protocorm (Fig. 11). Finally it passes 
out into the soil (Fig. 12). In no orchid studied in the present series (Odontoglossum, Oncidium, Cattleya, 
Cymbidium, Vanda, Cypripediutn, etc.) does the developing root when passing through the tissues enter the 
fungal zone nor do the hyphae extend into the root. In fact there is often a suggestion of a delimiting 
membrane separating the two areas {cf. Fig.12). Thus when the root enters the soil it is absolutely free from 
infection ; in none of the usually cultivated orchids does the root receive fungus from the swollen protocorm. 
Infection takes place from the soil most frequently when the root is about a quarter of an inch in length, 
the hyphje entering by the root hairs a little behind the region of greatest growth. This throwing off of the 
fungus, as it were, is repeated in orchids with tubers which do not retain their roots : the tuber is not infected 
and the new roots receive their fungus from the soil. In fact, in orchids so far studied it is only in the 
saprophytic Neottia that constant infection obtains. Here infection progresses gradually from the widely 
infected protocorm into the body of the plant, gains the rhizome and infects the successive roots. The 
region of infection is thus perfectly continuous throughout the plant from the tip of the protocorm to the 
base of the inflorescence : as Bernard remarks, according to the evidence the whole of the mycelium harboured 
by a Neottia has for its single origin the mycelial filament which first penetrates the embryo.^^ 

The question arises as to whether root infection per se is obligate in orchids with abundant chlorophyll 
or whether it is a necessary evil. If the latter, one would expect the fungus to be lodged in the roots, though 
restricted in distribution. As stated above, all the cells entered seem to act as digestive cells in cultivated 
orchids. Is such digestion a device for protection or for nutrition ? 

What has been happening to the fungus during these stages ? The course of events was first followed 
by Bernard. As we have seen, the fungus enters at the suspensor end of the seed by the cells of the suspensor 

^* The association can be even more close under certain conditions. Flower scapes are frequently unable to pierce the humus 
covering them and the flowers and seeds develop underground, sometimes beneath the root-tufts which produce them. Mycelium 
apparently from the rhizome of the plant passes up the centralcavity of the stem and infects the seeds in the subterranean fruits 
which are thus able to germinate. 

vii. 



near the point of attachment (Odontoglossum) or by the cells of the pole of the embryo where the suspensor 
is attached {Vandd). There appears to be an attraction, though feeble, towards the place of entry. The 
first filament entering the seed apparently excludes all others, though it may be of an attenuated form and 
unable itself to bring about germination. Bernard compared this with vaccination : the infection immunizes 
the seed. In successful germinations the fungus, after seed entry, follows the development of the cells 
forming mycelial balls in all the posterior portion of the seedling. According to Bernard, when the fungus 
reaches the cells bordering on the meristematic region digestion takes place. This is regarded as being 
analogous to phagocytosis such as occurs in animals where the white corpuscles of the blood attack, engulf 
and digest any invading micro-organisms : the cells in which the digestion takes place are the phagocytes. 

In general these may be regarded as definite cells often recognisable, even before infection on account 
of their nucleus sometimes becoming lobed. The balling of the fungus in the cells is compared with agglutina- 
tion, and the manner in which this occurs only in cells of the developing seedling which have achieved their 
growth is compared with cases of mortal infection where the balling is abandoned sooner or later and the 
fungus grows on in every direction and invades all the tissues indifferently. 

Digestion eventually takes place in all the more deeply lying cells, while the external layers act as host 
cells. The fungus can pass out of the protocorm by way of the hairs present on its surface. 

This application of the theory of phagocytosis is a most attractive one. Gallaud" first suggested the 
similarity of the function of the digestive cells and that of phagocytes, but it is to Bernard that we owe the 
working out in detail." Much investigation on the germinating seed is still needed. Bernard's account of 
the distribution of the phagocytes is not satisfactory. As the photomicrographs (Figs. lo, 12, 13) show it is 
not unusual for all the infected cells of the protocorm to be able to digest the fungus eventually. 

Germination Without Fungus. 

How far is it possible to replace the fungus by artificial conditions ? Bernard concluded from a con- 
sideration of the way in which the endophyte can act at a distance, i.e., bring about changes in cells to which 
it has not access, that there is a general modification of the physico-chemical properties of the sap which 
can reach all the tissues. He tried the effect of solutions of salep and saccharose of increasing concentrations 
on seeds of Bletilla, Cattleya and Lalia. In Bletilla where, as we have seen, germination takes place with the 
formation of slender seedlings in the absence of fungi, in high concentrations most of the seedlings showed 
thickened protocorms and short internodes comparable with fungus infected individuals. The seeds of 
Cattleya and Lcslia at low concentrations swell and become green. With higher concentrations development 
is always much slower and more irregular than with fungi, but one can obtain seedlings of quite normal 
appearance. As the concentrations increase the development is increasingly better, but more irregular : 
but there is an upper limit beyond which there is no germination. 

Thus it appears that augmentation of the culture medium can, in certain cases, supply the place of 
fungus action. In fact Bernard states that in the condition of his experiments it was more certain ajid 
easier to germinate certain seeds by the action of concentrated solutions than to have recourse to fungal 
infection. Germination was slow, but very regular, the protocorms had a normal appearance and die 
seedlings when fairly developed could be transplanted. Experiments showed that Rhiwctonia was able 
to increase the concentration of the solutions in which it grew and Bernard considered it probable that it 
acts similarly in orchid tissues and increases the degree of concentration of the sap. This problem of 
autonomous germination recalls to mind that of parthenogensis — the development of an ovum without the 
intervention of a spermatozoon. The egg possesses all the substances necessary for activation : the 
spermatozoon is an inciting cause of these reactions within the egg system on which development depends. 
Parthenogensis occurs naturally in certain groups, but it has been brought about experimentally in numerous 
cases where fertilization normally obtains.'' ^"^ '9 Apparently the first successful attempt was made by 
Tichomiroff in 1886, who stimulated the unfertilized ova of the silk moth to development by rubbing them 
between two pieces of cloth. Various methods have since been used such as treatment with fatty acids, 
certain salts such as barium chloride, lipoid solvents such as chloroform, hypertonic and hypotonic 
solutions, etc.'" 

" F. Gallaud. Etudes sur les mycorrhizes endotrophes. Rev. Gen. Bot. XVII. pp. 5 et passim. (1905). 

1' Bernard (191 1) also showed that the bulbs of Loroglossum contain a diffusible substance which has a fungicidal* effect 
on Rhixoctania. 

*^ F. R. Lillie. Problems of Fertilization. Univ. of Chicago Science Series. {1919). 

1' Y. Delage and M. Goldsmith. La parthteogenese naturelle et exp^rimentale. Paris. (1903). 

™ The only case in which parthenogensis has been induced in the entire vertebrate phylum is in the frog, where Bataillon in 
1 9 10, after years of vain attempts, finally succeeded by the exceedingly simple method of pricking the eggs with a fine needle. 
It is necessary that blood or tissue extract should be carried into the egg by the needle. This method hai been abundantly 
confirmed and tadpoles so obtained have been reared to maturity by Loeb and Bancroft. 

viii. 



Another significant similarity is that artificially activated eggs always show a marked slowness in their 
rate of development, even with the best methods, as compared with the fertilized eggs. This suggests, 
according to Lillie, some factor that has not yet been successfully imitated in any artificial way. Is it 
possible that in both cases accessory food factors (vitamines) may play a part ? In considering the case of 
seeds it might be pointed out that there are many instances of peculiar germination known in other phyla. 
Pinoy '* showed that spores of Myxomycetes such as Chondrioderma difforme do not germinate unless bacteria 
are present. Ferguson*' discovered that the only way in which she could germinate the spores of the 
common mushroom effectively was by having a little mycelium of the fungus present in the cultures and 
Servettaz" found that a species of Oospora activated the growth of the moss Phascum cuspidatum to a 
remarkable degree, though the favourable action was of short duration in the conditions of his experiments. 

Gastrodia. 

An unusual and interesting type of mycorrhiza occurs in Gastrodia elata", a non-chlorophyllous orchid 
widely spread throughout Japan, where it occurs mostly in woods under Quercus serrata and Q. glandulifera. 
The full-grown flowering tuber is oblong and slightly curved, attaining almost without exception a length of 
10-17 <^™' This tuberous rhizome is the whole vegetative part of the plant and consists essentially of 
parenchjrmatous cells. Multiplication usually takes place by the tuber. It produces long rhizomes from 
its apex or node, upon which stalked off-sets are developed. At the end of autumn the mother body and the 
pedicel of the off-set undergo degeneration, so that the daughter tubercles are set free. Unless the mother 
tuber has been infected with the necessary fungus the off-sets decrease in size with each successive generation, 
until they become so much reduced and deficient in food materials that they are incapable of further multi- 
plication. The fungus necessary for proper development is not a microscopic mould as in the other orchids 
studied, but Armillaria mellea, the well-known " honey fungus." This toad-stool is extremely common in 
our woods where it is a most destructive parasite, " indeed more trees die, in Europe at any rate, from attack 
by this fungus than through any other parasitic agent."*' The fructifications are found generally on or near 
stumps. If the earth beneath the toad-stool be dug up it will be found to contain one or more black strands, 
resembling bootlaces, which are attached to the base of the stem. These rhizomorphs, as they are called, 
consist of densely compacted fungus mycelium. Further, the mycelium in the wood of the tree itself is first 
felted and grows up through the cambium to a considerable height : when the tree is dead and the bark has 
become loosened the mycelium is transformed into a tangled mass of flattened rhizomorphs. Early 
mycologists considered that they were here dealing with three different species of fungus — the toad-stool 
{Agaricus melleus), the rhizomorph under the bark {Rhizomorpba subcorticalis) and the rhizomorph in the 
ground (Rhizomorpha subterranea). 

It is with the subterranean rhizomorph that we are here concerned. It forms a cylindrical, smooth, 
black strand, usually i to 1-5 mm. in thickness. Its peripheral portion, the so-called cortex, consists of 
compact, pseudoparenchymatous, brownish mycelium with a comparatively thick wall. The middle layer 
is composed of a bundle of large thin-walled mycelia with numerous septa. The inner cavity of the strand 
is traversed by a loose bundle of very fine longitudinal hyphse rich in protoplasmic contents. 

When the tuber of Gastrodia is attacked by the rhizomorph, infection is effected by a sucker-like branch 
of the strand which penetrates the cortical cell layers, partly compressing the underljdng cells and partly 
dissolving their walls. This mode of infection is, of course, quite different from the ordinary endophytic 
mycorrhizal type where infection is affected as a rule by a single hypha (cf. p. iii). It very much resembles 
the manner in which the parasitic Cuscuta attacks its hosts, the rhizomorph creeping over the surface of the 
tuber and giving off the infection branches at intervals. On entering the tuber the hyphae of the various 
portions of the strand essentially retain their structure. The infected area of the tuber may be divided into 
thfee regions, according to the structure of the cells and the nature of the hyphae contained within them. 
The external region is composed of two or three layers of cells which contain a densely entangled mass of 
comparatively thick-walled hyphae ; the middle region is similarly composed, except that the hyphae are 
generally thin-walled and of various breadths and often arranged as a pseudoparenchyma ; the innermost 

^ i. Pinoy. Role de bacteries dans le d^veloppement de certains Myxomycetes. Ann. Inst. Pasteur. XXI. pp. 632. 
(I907).» 

^ M. C. Ferguson. A preliminary study of the germination of the spores of Agaricus campestris and other Basidiomycetous 
fungi. U.S. Dept. Agric. Bureau of Plant Industry. Bull. No. 16 (1902). 

" C. Servattez. Recherches experimentales sur le developpement et la nutrition des mousses en miUeux sterilises. Ann. 
Sci. Nat. 9 ser. XVII. pp. 111-224 (1913). 

** S. Kusano. Gastrodia elata and its symbiotic association with Armillaria mellea. Journ. Coll. Agric Imp. Univ. Tokyo. 
IV. 1-66 (1911). 

" W. E. HUey. The fungal diseases of the common larch. Oxford (1919). 

ix. 



region has large cells each containing a few, slender, slightly curved hyphae. The three regions correspond to 
the zones in the rhizomorph. The hyphse of each region show characteristic alterations. They are per- 
manent in the first region ; in the second they undergo self-disorganization ; while in the third they are 
mostly consumed by the cells of the host. The mode of development of the fungus in the middle region 
simulates the ordinary clumping seen in most orchids, but the course of events is different in that the proto- 
plast is consumed by the hyphae before their collapse takes place. The destruction of the protoplast shows 
the parasitic properties of the hyphs. The cells of the inner regions are apparently metabolic centres of the 
orchid where the food materials are elaborated. The nucleus and cytoplasm undergo remarkable alterations, 
and secondary products appear indicating considerable activities. After the disappearance of the hyphse 
the nucleus resumes its original form and.structure, while the cytoplasm again becomes fibrous and vacuolate. 
Starch grains disappear from all the mycorrhizal cells, to reappear in the inner region with the cessation of 
metabolic activity. 

The association of tuber and rhizomorph takes place quite occasionally. If a tuber forms mycorrhiza 
it can give rise to a full grown off-set which remains dormant during the winter and develops the inflorescence 
axis in the following year : otherwise no flowers are produced. 

So far no results have been published as to the germinatiomof the seeds of Gastrodia. One would expect 
that fungal infection is necessary for seedling development, burwhether the fungus is a form like Rhizoctonia 
or whether there is some adaptation by which Armillaria becomes operative remains to be seen. In either 
case the facts will be of the greatest theoretical interest. ' 

The course of events in Gastrodia gives some support to the idea that the relation of fungus and orchid is 
primarily one of parasitism on the part of the former. At times the rhizomorph attacks tubers and destroys 
them in a manner similar to that in which it treats potato tubers. Usually, however, the fungus is kept 
well under control and its hyphse prevented from spreading beyond their apportioned region — and even so 
being absorbed by the orchid cells. It is difficult to see what benefit the fungus can gain under these con- 
ditions. The subterranean strands are apparently unable to obtain nutriment from the soil, their function 
in the usual life of the fungus being that of " runners." It would seem that Gastrodia has turned the 
attack of these into one of service for transmitting nutriment from the oak stumps to which the fungus is 
attached, for its own benefit : a colourless saprophyte unable to grow or to flower without the aid of one 
of the most destructive parasites known ! 

Number of Seeds and Distribution of Fungus. 

When one sees the dense masses of seedlings thriving in the culture flasks one contemplates as to the 
course of events under natural conditions. The enormous numbers of seeds which are usually produced in 
the capsules of orchids must have struck the most casual observer. " Not that such profusion is anything 
to boast of ; for the production of an almost infinite number or seeds or eggs, is undoubtedly a sign of 
lowness of organisation. That a plant, not being an annual, should escape extinction, chiefly by the pro- 
duction of a vast number of seeds or seedlings, shows a poverty of contrivance, or a want of some fitting 
protection against other dangers." Darwin"' estimated that in Cephalanthera grandiflora a single capsule 
contained 6,020 seeds and that, therefore, a plant with the usual four capsules would have 24,080 seeds. 
Similarly Orchis maculata had 6,200 seeds in a single capsule, and thus aplant having the not unusual number 
of thirty capsules would produce 186,300 seeds : " As this orchid is perennial, and cannot in most places 
be increasing, one seed alone of this large number yields a mature plant once in every few years." In order 
to retain the number of individuals of a species stationary it is only necessary that one mature plant should 
be produced during the period of growth of the parent — if more occur the species will tend to oust out all 
other species. " Linnaeus has calculated that if an annual plant produced only two seeds — and there is no 
plant so unproductive as this — and their seedlings next year produced two, and so on, then in twenty years 
there would be a million plants.. . . It would suffice to keep up the full number of a tree, which lived on 
an average for a thousand years, if a single seed were produced once in a thousand years, supposing that this 
seed were never destroyed, and could be ensured to germinate in a fitting place."*' To give an idea of what 
the above figures for Orchis maculata really mean Darwin worked out the possible rate of increase. " An 
acre of land would hold 174,240 plants, each having a space of six inches square, and this would be just 
sufficient for their growth ; so that, making the fair allowance of 400 bad seeds in each capsule, an acre 
would be thickly clothed by the progeny of a single plant. At the same rate of increase, the grand- 
children would cover a space slightly exceeding the Isle of Anglesea ; and the great grandchildren of a 
single plant would nearly (in the rate of 47 to 50) clothe with a uniform green carpet the entire 

" C. Darwin. Fertilisation of Orchids. (1862). 
*' C. Darwin. Origin of Species. (1859). 

X. 



swface of the land throughout the globe " — and as 0. m»culata is perennial, the parent plant would still 
be alive ! 

But even these numbers in our native orchids are much exceeded by those of tropical species. Scott 
estimated that a capsule of Acrofera contains 371,250 seeds and, judging from the number of flowers borne 
by the plant, the total number of seeds for an indivivual would be 74,000,000 : Charlesworth estimated 
825,000 seeds for a single capsule of Cymbidium Traceyanum : MuUer 1,756,440 seeds for a single capsule of 
Maxillaria. It appears to be a general biological rule that where the conditions of successful germination 
are difficult of attainment a prolific number of seeds (or spores) are produced and vice versa, where the 
requirements are not of a specialized nature, a smaller number occur. 

In the case of orchids it seems not unlikely that the enormous seed production is in some way related 
to the fungus question. Their small size, their lightness, their net-work integument and the presence in 
some genera of elaters ensure their effective dissemination. But unless the necessary fungus be to hand no 
germination occurs — the seed may develop to a certain extent, but it does not produce roots unless the 
appropriate fungus enters its cells. 

So far, however, we know nothing of the distribution of these fungi in nature except so far as they occur 
associated with rooted orchid plants. Probably most people are aware that fungi of all kinds are present 
in the soil, but few realize in what enormous numbers they occur and the manner in which some are restricted 
to the soil. Hagem" calculated that in a gram of soil from a potato field, 350 spores of Rhizopus stolonifer 
and 250 each of Mucor sphtsrosporiis, M. nodosus, Absidia cylindrospora and Zygorhynchus Moelleri were 
present ; and these numbers are much exceeded by Penicillium (90-95 per cent, of spores in uncultivated 
soil according to Sopp*») and other Hyphomycetes. Traaen'" calculated that from 10,000 to 120,000 spores 
of Geomyces vulgaris and from 1,000 to 20,000 spores of Humicola fuscoatra occur in a gram of soil. Much 
work has been done recently on the biological activities of such fungi, attention being paid chiefly to cellulose 
destruction and the possibility of nitrogen fixation. It is extremely probable that certain of the forms 
isolated are capable of acting as mycorrhizal fungi, though none have apparently been recognised as such. 
Further it is possible to isolate Rhizoctonia from the soil in the immediate neighbourhood of orchid plants 
growing wild (as also from the soil of pots containing cultivated orchids) : but notwithstanding the large 
number of species of soil fungi isolated it does not appear to have been found, or at least recognised, by any 
investigator. We are thus lacking in data as to the distribution of orchid fungi in the soil. Since, however, 
Bernard isolated Rhizoctonia repens from many European orchids and showed it to be the commonest 
endophyte amongst cultivated species, it must be of world-wide distribution, since in order to account for 
the distribution of the orchids it is necessary to assume that this particular fungus must occur practically 
wherever orchids grow. 

■J Ericace.^. 

i 

A family of plants which is usually linked with orchids as showing the same constancy of fungal infection 
is the Ericaceae. Frank early realized that the relation between the fungus and flowering plant in these two 
families is a particularly close one. In certain ericaceous plants he remarked on the absence of root-hairs, 
the absence of, or reduction in, the amount of cortical tissues, the reduction of the root-cap, and the masses 
of fungus mycelium in the enlarged cells of the epidermal layer. Ternetz'^ was successful in isolating the 
fungi from certain species and growing them in pure culture, constantly obtaining the same fungus from 
the same species of flowering plant. All the fungi belonged to genus Phoma" — one of the Fungi Imperfecti, 
but o.. a totally different group than is Rhizoctonia — and were apparently morphologically and physiologi- 
cally distinct. She showed that infection of Calluna took place in the seedling and also found infection in a 
case of viviparous germination in Andromeda. 

Rayner" working with Calluna vulgaris was able to show that the full development of the seedling was 
dependent upon the presence of the mycorrhizal fungus — there is here an " obligate symbiosis " of a type 
very similar to that in orchids. Finding that the sterile seedlings were unable to form a root-system she 
investigated the matter in the manner made classical by Bernard. The seed coats were found to become 
infected while the seeds are still in the ovary. Delicate branched hyphae are present in the cells of the ovary 
wall, in the tissue of the central column and in the funicles of the seeds. Branches of this mycelium grow 

" O. Hagem. Untersuchungen uber Norwegische Mucorineen II. Skrifter Vidensk-Selsk. Christiania. I. Math.-Natur. 
Kl. No. 4 (1910). 

^^ O. J. O. Sopp. Monographie der Pilzgruppe Penicillium. idem. No. 11 (1912). 

•" A. E. Traaen. Untersuchungen uber Bodenpilze aus Norwegen. Nyt. Mag. Natunvidensk. LII. pp. 19-121 (1914). 

" C. Ternetz. Uber die Assimilation des atmospharischen Stickstoffs durch Pilze. Jahr. f. wissensch, Bot. XLIV. 
pp. 353-408 (1907). 

" Phmna radicis-Oxycocci, P. radicis-Andromeda, P. radicis-Vaccinii, P. radicis-l elralicis and P. radicis-Erica. 

'* M. C. Rayner. Obligate symbiosis in Calluna vulgaris. Ann. Bot. XXIX. pp. 97-133 (1915). 

xi. 



across from the cells of the ovary wall to those of the seed-coats, extending from one seed to another. The 
fungus was isolated and grown in pure culture. It proved to be a pycnidial form similar in all respects to 
the genus Phoma. Sterile seeds sown on this develop normally, whereas in its absence the seedlings merely 
form a few reddish or chlorotic leaves, but no roots. Infection of the seedling root takes . place at, or 
immediately after, it emerges and may begin at the tip by hyphae forcing their way between the cells of the 
apex, though more usually it occurs simultaneously at several points. The mycelium immediately becomes 
intercellular and infection spreads rapidly from cell to cell. Some hyphal branches grow out and infect 
fresh rootlets as they develop ; others form a tangled skein of fine hyphas in the superficial cells. One of 
the most interesting points of the story is,however, that the fungus does not remain confined to the roots but 
infects the whole of the young seedling. In the subaerial parts the mycelium does not develop so extensively 
on the surface of the plant, nor do the hyphae become balled up in the superficial cells as in the roots, but are 
irregularly distriblited in the tissues. In the mature plant likewise the fungus is not confined to the roots 
but is present in the tissues of the stem, leaf, flower and fruit. The hyphs can also be seen ramifying among 
the hairs or closely appUed to the cuticle of the epidermal cells : they show no preference for special points 
of entrance or egress, penetrating with equal ease the cuticularized cells of the epidermis or the base of a hair. 
The ovary — and later the young fruit — contains mycelium in all parts of the internal tissues. This mycelium 
infects the seed coats of the developing seeds. The embryo and endosperm of the resting seed are free from 
infection. 

Thus, as in Neottia, we are dealing, except in the seed, with a dual organism. The type of association 
is, however, different from what obtains in the orchids so far studied, where no such distribution has been 
found — and an analogous constancy apparently only occurs in non-chlorophyllous genera. From the fact 
that Rayner has recorded the presence of ovarial infection in a number of Ericaceae — Rhododendroideae, 
Arbutoideae, Vaccinioideae and Ericoideae — it may be that the fungus is similarly distributed throughout 
the tissues of these plants, and presumably obligate symbiosis is to be inferred. 

In no other case has the necessity of the presence of the mycorrhizal fungus for germination been proved. 
There can be hardly any doubt, however, that such a phenomenon is not restricted to two groups so widely 
separated as the Orchidaceae and the Ericaceae. What have these families in common ? Apart from the 
similarity in habitat of certain species there seems to be nothing except the smallness of their seeds — and it 
is naturally to seed characters that one looks in this connection. As we have seen, the seeds of orchids are 
exceedingly small ; reduction in most genera would appear to have reached its limit. In typical Ericacex 
the seed is very small, rarely exceeding 2 mm. and often less than half this size. There is a richly developed 
endosperm in which a straight embryo is embedded one-half to two-thirds the length of the seed, always 
showing a root, an axis and two cotyledons more or less differentiated. It is also of interest to remark that 
such genera as Kalmia and Ledum have a net-work integument to the seed. 

PyROLACE/E. 

Allied to the Ericaceae is the family Pyrolaceas writh the sub-families Pyroloideae and Monotropoideae. 
In families of flowering plants which show saprophytism and parasitism there usually occur green purely 
autophytic plants, with typical green leaves and numerous flowers ; plants that are purely saprophytic or 
parasitic, with colourless scales and a reduced number of flowers ; and all gradations between. Henderson'* 
instances the families Burmanniaceae, Orchidaceae, Gentianaceae and Ericaceae as examples of this. Regard- 
ing the Pyrolaceae as a saprophytic sub-family of the Ericaceae we can trace a relation between increasing 
saprophytism and a more intensive development of mycorrhiza. In the root tip region we get an ascending 
series in the amount of fungus present from Chimaphila umbellata where the epidermal cells of some roots 
are without hyphae and other roots with hyphae, but not in every cell, to C. maculata with a greater number 
of the epidermal cells filled with hyphae ; in Pyrola rotundifolia and P. elliptica all the cells are infected, 
and there is the beginning of intertwined hyphae round the root tip ; then in Monotropa Hypopitys an 
increase in the width and extent of the sheaths and a division into two zones — an outer loosely woven mass 
of hyphae and an inner more compact one — and finally in M. unifiora a still greater width of the fungal 
sheath. In the least saprophytic species the epidermis soon dies off, carrying with it the fungal hyphae as 
in Chimaphila and Pyrola, whereas in Mowoiro^a, especially M. uniflora,the epidermis is still living and filled 
with hyphae when the root is quite old. 

Corresponding with this increase in saprophytism there is an increase in the number of seeds produced 
and a reduction in their size and structure. "The endosperm in the Pyrolaceae consists of relatively few large 
cells — the embryo of about twenty-five to thirty cells with no trace of cotyledons. In the Monotropaceae 

^* M. W. Henderson. A comparative study of the structure and saprophytism of the Pyrolacese and Monotropacex with 
reference to their derivation from the Ericaceae. Contrib. Bot. Lab. Univ. Pennsylvania. V. pp. 42-109 (1919). 

xii. 



the number of endosperm cells is still less and the cells are larger, the embryo also is very small, composed 
of only nine or five cells." As these seeds also have their integument in the form of a net-work there is an 
exceedingly close superficial resemblance to those of orchids. 

Comparing the members of the Pyrolaceas as a whole with the Ericacese it would seem exceedingly 
probable that their seeds are even more dependent upon infection by the mycorrhizal fungus than are those 
of their chlorophyllous relatives. It will be interesting to learn at what stage infection takes place and 
whether or not a close approximation to the more advanced orchid type obtains. It is probable that the 
fungus will be found to be generally distributed in these plants as in Calluna. 

BURMANNIACE/E AND GeNTIANACE^. 

The other two families in which mycorrhizas are typically developed are the Burmanniaceae and the 
Gentianacese" — in fact Stahl considered that from this point of view the latter family areas important as the 
Orchidaceas. Moreover, in these families the seeds are small and numerous, with little reserve food material 
and no chlorophyll. Further there are the typical gradations from green plants to colourless saprophytes 
and correlated with this is an increase in number and decrease in size of the seeds, with a change in the 
embryo until we end in the most reduced examples with little differentiated or formless masses, and an 
increasing amount of fungus in the roots. The seeds of the saprophytic genera have a network integument 
and in appearance bear a very close resemblance to those of orchids. The Burmanniaceae are closely related 
to the Orchidaceae, and we should expect that showing so many characters in common there would also be a 
resemblance in the important one of obligate fungal infection for germination. In the Gentianaceae there 
are many isolated records of difficulties in obtaining seed germination in some of the genera, and it is common 
knowledge that many Gentians are difficult to raise from seed. It would seem extremely probable that in 
this family also the mycorrhizal fungus is necessary for seedling development. 

Ceillier" has worked out in detail the relation between the presence of mycorrhiza and small seeds. 
In certain cases as in Juncaceje the seeds are small and little differentiated, but as they possess chlorophyll 
they are able to begin photosynthesis immediately on sowing. Small seeds with much reduced embryos, 
also occur in parasitic forms such as Cuscuta, Orobanche, etc. No fungus is present in these genera, but 
apparently germination is not successful unless contact is made with the organs of the requisite host. It 
may be that the stimulus necessary in these cases is analogous to that requisite to bring about root formation 
in plants with obligate mycorrhizas. 

Origin of Saprophytism. 

What is the trend of evolution in plants of which the roots are normally infected with endophytic fungi ? 
A general survey of families in which endotrophic mycorrhizas are typically developed shows that it is the 
rule for these families to have small seeds ill-adapted for successful germination. It has also been proved 
for orchids and for Calluna that the seeds need to be infected by the mycorrhizal fungus before the seedling 
can produce roots. Further it is in these families that typical saprophytic species occur (if we concede that 
the Pyrolaceae are saprophytic Ericaceae) : in fact the presence of fungi in the roots of saprophytes is so 
common (the apparent exception being Wullschltegdia), that MacDougal regards these seed-plants as being 
" saprophytic symbionts." *' Without the necessary data it is doubly unsafe to theorise, but it suggests 
itself that in families adapted to a mycorrhizal habit there is a tendency for the seed to become dependent 
upon the fungus for successful germination, and there is a correspondingly greater production of seed. It 
has been customary to associate increasing saprophytism with the greater development of mycorrhizal fungus. 
May it not be rather that saprophytism has arisen by the mycorrhizal fungus taking over some of the 
functions necessary in germination and relieving the flowering plant of the need of excessive food production 
for the developing seed and thus of the necessity for carbon assimilation ? (The great amount of fungus 
in the roots of saprophytes militates against the idea that the root may be simply a lodging place for the 
fungus to be at hand for germination and of no use in nutrition). We see in Calluna an almost perfect device 
for the infection of the seed, and the fungus is generally distributed. The most general infection so far 

^^ " Most of the Orchideae are humus-plants, and it is noteworthy that dicotylous saprophytes, such as the Pyrolaceae, the 
gentianaceous Voyria, and others, show a reduction of the embryo like that of the Orchidea;. In Monotropa the embryo has but 
nine cells. The germination of the seeds of these dicotylous saprophytes is unknown. It takes place only in the presence of 
very special surroundings. Probably the fungi which are found in the roots in symbiosis are essential. The smallness of the 
seeds allows of a large number being formed, and thus the probability that one of the seeds at least will reach favourable con- 
ditions for germination is increased." Goebel, Organography of Plants. Part II. pp. 254. [1898] 1905. 

"R. CeilUer. Recherches sur les facteurs de la repartition et sur le role des mycorrhizes. Those. Paris (1912). 

"Johow (i889)places all the known saprophytic flowering plants in the six families Orchidaceae, Burmanniacese,Triuridaceie, 
Pirolea;, Monotropea and Gentianaceae. (The Triuridacese are a small family of tropical saprophytes with the two genera Sciaphila 
and Triuris and about forty species). 

xiii. 



found in orchids is in Neottia, which, as has been pointed out above, is most comparable with Calluna. But 
Neottia is saprophytic. In chlorophyllous orchids it almost looks as if when the necessary stimulus is given 
for seed germination precautions are taken to prevent general infection, the primary root even being free. 
In orchids digestion of the endophyte may also be a means of preventing general infection (though in 
Neottia this property can be easily recognised). Does such a general infection as we get in Calluna ultimately 
lead to saprophytism of the type seen in the Pyrolaceae ? Are the events described above in the germination 
of certain orchids an effort to prevent general invasion and the " perfect symbiosis " of Neottia ? 

LOLIUM. 

A case which recalls to mind that of mycorrhiza — especially having regard to recent discoveries — is that 
of the grass Lolium. The fact that the grains of Lolium temulentum contain a layer of fungal hyphae situated 
between the aleurone layer and the fruit and seed coat was first demonstrated by Vogl in 1898, and since 
then has been many times investigated in different species of the genus. The latest worker is McLennan" 
who used Lolium perenne for her researches. The fungus is far more common in the genus than has hitherto 
been thought, and it is remarkably constant. Every seed examined (169 of L. temulentum and 115 of 
L. perenne) showed infection. The fungus is endophytic, occurring within the cells. It is present in the 
embryo sac at, or immediately after, fertilization : thus there is a material difference from what happens in 
orchids and Calluna. The fungus increases in quantity at the expense of the nucellus and the cells of the 
carpel wall. As the endosperm is formed the fungus is absorbed as a source of food supply for the developing 
embryo. The ovum is infected before any divisions have taken place in it. 

The hyphx already in the very young embryo, follow the development of the stem-apex and remain 
localised in their growth until germination takes place. The growth of the fungus keeps pace with that of 
the plant : the hyphae, however, are mainly restricted to the growing apex, but can be seen extending for a 
short distance down the stem. Even at this stage the intracellular nature of the fungus can be demon- 
strated. Some of the parenchymatous cells of the grass are invaded and used as a food supply by the 
hyphae. When the inflorescence is formed the fungus is especially abundant at the base of the carpels. 
The cells so affected do not increase in size, and are only to be distinguished from normal unaffected cells 
by their different staining properties. It is not till the ovule is well advanced that any great increase in the 
fungal partner takes place. The fungus has not yet been isolated." It has been suggested that it is probably 
a degenerate member of the Ustilagineae (Smuts) or of the ergot type. The former would seem the more 
likely. Smuts attack grasses very generally and often it is the flower that is infected and later the seed, and 
thus the whole plant. On general grounds it would appear that the line of development to the stage found 
would be the gradual subjection of a parasitic fungus such as Ustilago rather than the further development of 
a typical mycorrhiza. An examination of Lolium roots shows that no typical endophytic fungus is present — 
in fact these are peculiarly absent in the Gramineae, though recorded by Schlicht for Holcus lanatus and 
Festuca ovina and by Tubeuf for certain moorland grasses — and the area of infection seems limited to 
the region of the stem apex. Thus, though it would appear at first sight that the progress of evolution 
had been along a line similar to the Calluna type leading to infection of the embryo as apart from the 
seed-coat, and consequent continuous infection, it is more likely that in the typically non-mycorrhizal 
grasses such a union has been brought about by a subjection of a seed parasite. 

Relation between Fungus and Flowering Plant. 

Throughout the preceding pages incidental remarks have been made regarding the relation between 
the two constituents of the mycorrhizal association. The subject is one of extraordinary interest and of 
extreme difficulty. It does not seem possible to regard all such associations as being of the same nature or 
as having arisen in the same way. 

As we have seen Rylands was the first to record fungi in association with roots, though his account is 
not very clear : his idea that the fungus performs no essential function in the economy of Monotropa is one 
that has had few supporters. 

Reissek, who in many ways seemed before his time in his attitude towards the subjecvregarded the 
regularity and permanence of the presence of fungi in orchid roots as of great importance. He apparently-. 
considered that they were not absolutely necessary for the life of the plant and suggested that the orchid 
could generate without the root fungus in the same way that the greater number of flowering plants are able 
to propagate without flowers. 

Tlie gradual realization of the dual nature of lichens brought in its train the conception of symbiosis 

" E. McLennan. The endophytic fungus of Lolium. Part I. Proc. Roy. Soc. Victoria XXXII (^f.S.) pp. 252T301 (1920) 
"• Fuchs (Hedwigia LI., pp. 221-239 (191 1)) claims to hare proved that the fungus is a species of Fusarium. 

xiv. 



but the increasing knowledge as to the nature of fungus-roots played a not inconsiderable part in the growth 
of the idea. 

From the year 1862 Tulasne began to consider the relation between the False Truffle (Elaphomyces) 
and the roots of trees as one not of simple parasitism as he had previously (1841) thought, but one by which 
both organisms benefited in some way. Pfeffer in 1877 took up this idea of mutual benefit and made it 
more precise. Other workers — Treub, Goebel, Kamienski — also regarded the relation between fungus and 
root as of this description, .^t is to the work of Frank, beginning in 1885, that we owe a proper conception 
of the widespread phenomenon and a clearly outlined theory of symbiosis between fungus and root. Natu- 
rally as more facts both of observation and experiment were obtained Frank's original theory was somewhat 
modified— originally it was that plants with ectotrophic mycorrhiza did not themselves draw nutriment 
from the soil, but that the mycelial filaments which completely envelop the absorbent roots procure for it 
all its nutriment. Such roots always lack absorbent root hairs. The absence of these organs of absorption 
corresponding to the presence of mycelial filaments suggests that the latter take up the functions of the 
former. Later, the view taken was that the fungus does not necessarily nourish the roots, but draws its 
nutriment from the humus of the soil and passes on a portion of this to the roots. In other words the presence 
of the fungus allows the root to make use of certain substances of the humus that it would be incapable of 
utilizing in its absence. Another hypothesis which figures largely in the literature of the subject is that of 
Stahl*" This author endeavours to show that the role of the fungus consists in furnishing the plant with 
mineral nutriment. Comparing plants with and without mycorrhizas he points out certain diflterences 
which always appear to indicate a much greater circulation of water in the latter. Thus their roots are 
strongly developed, they possess numerous root hairs, their leaves transpire energetically and are often 
provided with water stomata. Further, their tissues are ordinarily rich in starchy matters and poor in 
sugar, i.e., in a condition favourable for transpiration. The fact that mycotrophic plants transpire less" 
and are in consequence less well fed in nutrient soils leads to the idea that the service which the fungus 
renders to the host consists in remedying the insufficiency of transpiration. Stahl imagines that the fungus 
hands over the products of assimilation of the salts rather than the salts themselves. There exists between 
phanerogams and fungi growing in the humus of forests, heaths, moors, etc., a competition for the salts 
which the vegetable debris already contains in a concentrated form. The advantage in this struggle 
would apparently be on the side of the fungi owing to their mode of life. Plants with very active trans- 
piration are alone capable of struggling with success against fungi in soils rich in humus : plants with feeble 
transpiration are only able to subsist in these conditions by the help which their symbiotic fungus brings. 
Magnus (1900) from his anatomical investigations regarded the digesting cells as serving for absorbing 
the nutriment of the fungus : the lodging cells, on the other hand, are set apart for the nourishment of the 
fungus on the cell contents and for its hibernation. This idea would give the classical balance of 
symbiosis — each component benefiting to an approximately equal degree. 

Gallaud regards the communication of the endophyte with the exterior in endophytic mycorrhizas 
as insufficient to assure to the plant the absorption of nutritive substances. From a study of numerous 
types of infection he holds that the fungus when in the root leads a life independent of the exterior and 
that it must therefore obtain all its nutriment from the plant. Comparing its mode of life with that of 
fungal parasites such as Peronosporaceae he decides against its parasitic nature and regards it as a special 
form of saprophyte — an internal saprophyte. 

Ternetz working with the fungi from Ericaceae records as a result of careful experiments that they are 
able to fix free nitrogen. From a theoretical point of view this is of extreme interest fitting in well with 
what is known concerning the bacteria in the root nodules of the Leguminosa, but so many discordant 
results have been recorded in such studies that it would be well not to accept these without confirmation. 
Incidentally it may be again remarked that Burgeff was unable to show any such fixation in orchids. 

Owing to the totally different complexion that Bernard's work put upon the mycorrhiza question 
his views are of particular interest. He regards the fungus in orchids as a parasite : an orchid 
suffers from a benign cryptogamic malady. Symbiosis for him represents the immunity realized by 
phagocytosis. 

Burgeff on theoretical grounds considers that both orchid and fungus must benefit by increased power 
of reproduction. He is in general agreement with Stahl as to the nature of the benefit the flowering plant 
receives. The union arose originally from the ability of the fungus to take up carbon compounds from the 
soil. The function of the fungus in germination is to introduce a solution of carbohydrates into the seed 
by means of its enzymes. 



*°E. Stahl. Der Sinn der MycorrhizenbUdung. Jahr. f. wissen8ch Bot. XXXIV. pp. 539-668. (1900). 
" The difficulty in drying orchid plants for herbarium purposes is a result of this. 



XV. 



Most recent workers on ectotrophic mycorrhizas regard the fungus as parasitic. Fuchs*' attempted 
to inoculate the roots of Abietineae by adding fungus spores to the soil. He did not succeed in his ex- 
periments, but regarded the vehemence with which the young plants cut off the infected cells as an effort 
to prevent the attacks of a parasite. 

Weyland" introduced the microchemical method of studying the question and it is probable that 
from such studies a clearer idea of what is really taking place will be obtainable, by the determination of 
the localization of nutriment. He considers that the fungus in an ectotrophic mycorrhiza is really a parasite 
and has nothing to do with symbiosis. 

Weevers'* working from a chemical point of view on the presence of ammonia and ammonium salts 
in plants established the fact that although ammonium salts were found in abundance in the tubercles of 
the Leguminosae they were in small quantity or absent in mycorrhizal plants. He holds therefore that it 
fungus-roots really assimilate nitrogen it must be brought about in a manner different from that in the 
Leguminosae. Weevers is rather of the opinion that mycotrophic plants are, with the help of their fungus 
partner, able to utilize fully the organic compounds of the soil. 

McDougall,'* working with ectotrophic mycorrhizas of forest trees formed by the association of toad- 
stools with the roots, considers that they are not in any sense symbiotic associations but must be considered 
as instances of parasitism by the fungi. 

Rexhausen *" studied ectotrophic mycorrhizas by the microchemical method. He considers that the 
fungus and the root together form an osmotic unit for the absorption of nutrient salts. These are probably 
made soluble for the root by the fungus. This gathering up of nutrient salts is first used by the fungus 
for its own benefit. The mycorrhiza is not a fixed symbiotic condition, but is dependent upon the biological 
condition of the soil. Where the conditions are not suitable for the growth of the fungus it acts as a 
parasite on the root and may damage it severely, as it cannot be kept in check. Where the fungus is well 
nourished it can be easily withstood by the root. Thus in good soils the mycorrhiza gradually disappears 
or, at all events, the fungus part becomes less. 

It will be apparent from the above that many somewhat diverse theories have been put forward to 
account for the fungus-root association and many modifications have been proposed. No purpose would 
be served here by entering on a detailed criticism : the only general one we would suggest is that no benefit 
can result from pushing the old idea of mutual and equal advantages of the two components to its extreme. 
Referring only to orchids it seems most reasonable to regard the condition as having arisen from parasitic 
attacks by the fungus. This seems beyond doubt in the exceptional case of ATmillaria and Gastrodia. 
The ability of the fungus to transport nutrient solutions has been made use of by the flowering plant. 
As in the case of Leguminosae and their nodules the tables have been turned and the " host" has become 
the aggressor, even attracting the fungus to the embryo. We are short of definite facts — there is a 
conflicting mass of detail on such an important point as the relation between the endophyte and the 
soil — and until these are obtained one theory seems as good as another. 

It would be indeed strange if the difference between ectotrophic and endotrophic mycorrhiza should 
resolve itself into a case of the fungus being parasitic on the flowering plant in the former, while in the latter 
the flowering plant is parasitic on the fungus. 

I am indebted to Mr. E. H. Ellis for the photomicrographs, with the exception of Figs. I and 4, for 
which I must thank Mr. R. J. Tabor. 



While the above was in the press an important paper by H. Christoph entitled " Untersuchungen 
uber die mykotrophen Verhaltnisse der ' Ericales' und die Keimung von Pirolaceen " appeared in Beih. 
Bot. Centralbl. XXXVIII. pp. 1 15-157 (1921). In it the author controverts the results obtained by Rayner 
concerning the necessity of the root-fungus for seed-germination (cf. p. xi). It should be noted, however, 
that he has not seen the full description of Dr. Rayner's researches, but apparently only an abstract of 
her prehminary account. Christoph concerned himself with the manner in which the fungus reaches the 
roots of the Ericaceae whether from the soil or from the seed. His first series of experiments were per- 
formed with cuttings. He took both large and small green side shoots from plants of Calluna vulgaris 

*^ J. Fuchs. Ueber die Beziehungen von Agaiicineen und anderen humusbewohnenden Pilzen zur Mycprhizenbildung der 
Waldbaume. Bibliotheca Botanica LXXVI. (191 1). 

*'H. Weyland. Zur Emahrungsphyioldgie mykotroper Pflanzen. Jahr. f. wissensch. Bot. LI. pp. 1-80 (1912). 

**T. Weevers. Das Vorkommen des Ammoniaks und der Ammonsalze in den Pflanzen. Receuil des Traveaux botaniques 
Nfcrlandais. XIII. pp. 63-104. (1916). 

**W. B. McDougall. On the mycorrhizas of forest trees. American Journ. Bot. I. pp. 51-74 (1914). 

*'L. Rexhausen. Uber die Bedeutung der ektotrophen Mykorrhiza fur die hoheren Pflanzen. Beit. i. Biol, der Pflanzen. 
XIV. pp. 19-58 (19")- 

xvi. 



both wild and cultivated. These were planted in shallow pots in humus heath soil — the soil in the one 
pot being sterilized and that in the other not. In both experiments a number of cuttings struck and suc- 
ceeded in establishing themselves. The roots of the cuttings in unsterilized soil became slightly infected, 
but no fungus could be found in those growing in sterilized soil. On replanting and transferring the latter 
cuttings to sandy soil they still remained free from fungal infection and continued in that condition for 
two and a half years. 

Similar experiments with cuttings of Erica camea gave analogous results. Both series succeeded 
and those planted in sterile black heath soil, and after one and a half years transferred, remained free from 
fungus infection for two and a half years. 

The plants without fungi in their roots were in just as good a condition as those which became infected 
and Christoph is of the opinion that the fungus is of no assistance to the plants and must be regarded as a 
harmless parasite. 

A second part of the paper deals with germination experiments with these two species. The results 
of thirteen experiments are summarized, though the complete account is not published 

Different soils were tried, both sterilized and unsterilized. Seeds of Calluna and Erica were sown in 
these, some having their coats sterihzed, some being used just as they were taken from the capsules. The 
results were similar in both series of experiments, except that Erica camea germinated only in the absence 
of light. Germination occurred in all experiments, e.g., sterilized seeds germinated in sterilized soil. Only 
those seedlings growing in unsterilized soil become infected with fungus whether the seeds are sterilized 
previously or not : in certain cases seeds which were taken from capsules in which a fungus was very obvious 
did not give rise to infected seedlings when sown in sterilized soil. The author concludes that infection 
of the root always comes from the soil and never from the seed coat. 

Regarding infection in the capsule, Christoph states that so long as the carpels are still green and the 
seed white a fungal infection of the tissue can never be observed. 

The author succeeded in extracting the fungus from the roots of the plants, but was unable to obtain 
spores in pure culture and was therefore unable to identify it. That it was probably the appropriate fungus 
was shown by infecting seedlings of both Calluna and Erica. 

The Ericales are considered to be facultative mycotrophic plants, since specimens growing in normal 
conditions always have fungus in their roots. In very dry places, however, plants of Calluna vulgaris and 
of Erica camea are often without fungi ; and in pot cultures allowed to become dry the fungus soon dis- 
appears. 

The third portion of the paper deals with the Pyrolaces. Working with Pyrola uniflora, P. secunda, 
P. minor and P. rotundifolia it was found that the hyphje of the infecting fungi had clamp-connections 
(and were therefore probably Basidiomycetes). The conclusion reached is that here also no true " sym- 
biosis " exists — infection depends upon many external factors, of which temperature, soU, moisture and 
aeration are the chief. Coralloid roots are not brought about by infection : there is a special development 
of the large epidermal cells and these, owing to their function of absorption, are specially suited for fungal 
development. 

In Monotropa the fungus possesses no clamp connections. 

The author was successful in germinating seeds of Pyrola rotundifolia which he chose, as they were the 
largest of the four species. The best results were obtained from : — i. Strong concentrated soil-extract ; 
2. Addition of peptone solution ; 3. Sowing on humus from habitat of plant — on sterilized soil there 
was no germination ; 4. Keeping cultures in the dark ; 5. Moderate moisture. 

It is suggested that the carbon compounds of the highly concentrated soU extract, acting in com- 
bination with the peptone, brought about germination by chemical action. 

Parallel experiments with peptone solution alone, soil extract alone, and with a mixture of both gave 
a slight germination in peptone solution, a stronger one in soil extract, but much the best is a mixture 
of the two. 

With regard to the question of infection of the seedling root from the capsule it is obvious that there 
is great discrepancy between the accounts of Rayner and Christoph, and until the results of one or other 
worker be confirmed it is not possible to draw from thein theoretical conclusions. That cuttings of Calluna 
can strike and come to maturity in sterilized soil without root infection is somewhat unexpected on account 
of Rayner's clear description of the distribution of the fungus in the plant ; in cultivated orchids it is 
quite likely that after the seedling stage fungal infection is not necessary. 

Concerning the germination of Pyrola rotundifolia seeds the account is not fuU enough to draw from it 
any theoretical conclusions. The fact that the seeds can be brought to germinate by chemical means is 
not surprising : it is analogous to what has been found by Bernard in Cattleya. There was apparently no 
attempt made to try the effect of the root-fungut on germination. 

December, 1 9a I. 

xvii. 



FIGURES. 



1. Transverse section of root of Habenaria 
just above the root tip. The dark 
masses show where digestion of the 
fungus is taking place. X 36 

2. Seed of Cymbidium, stained and 
mounted whole. The embryo is seen 
as an oval black patch within the net- 
work integument. X 56 

3. Longitudinal section of a seed of 
Odentoglessum. The anterior end shows 
smaller cells, the posterior end larger 
cells. (The integument has been rup- 
tured in making the preparation.) X2IJ 

4. Fungus from Odonieglossum (fihizoc- 
tonia lanuginosa Bern.) at the beginning 
of sderotium formation. X 36 

5. The same more highly magnified show- 
ing chains of " spores." x 215 

6. Seed of Odontoglossum sown seven days 
on a culture of the fungus : stained and 
mounted whole. X 56 

7. The same more highly magnified, x 215 

8. Longitudinal section of a seed of 
Odontoglossum nine days after sowing. 
The fungus has entered the larger cells 
at the suspensor end of the seed and 
formed balls of hyphs. (The integu- 
ment has been broken in cutting the 
section, cf. Fig. 6.) X 215 



9. Section of protocorm of Odontoglossum. 
The growing point of the stem can be 
seen at the upper end and the first and 
second leaves (Section not quite median). 
The fungus in many of the cells is 
already digested. X 56 

10. Later stage of Odontoglossum showing 
the beginning of the formation of the 
central stele and root. The stem 
growing-point is well developed. The 
fungus in the cells is mostly digested. 

X36 

11. Protocorm of Vanda showing the young 
root absorbing its way out of the side of 
the protocorm away from the fungal zone 

xig 

12. Odontoglossum seedling after the forma- 
tion of the first root. The root is not 
infected from the protocorm. (The 
dark patches in the root are raphides). 
X36 

13. Cells from fungal zone of Fig. 12 more 
highly magnified. The fungus is 
" clumped " (c/. Fig. 8). A fungal 
hypha is seen passing through the 
cell-wall. X215 

14. Longitudinal section of the root of 
Habenaria near the tip. Digestion is 
more prominent in the older (upper) 
portion of the root. X 18 

15. Longitudinal section of an aerial root 
of Epidtndrum showing infected cells 
in the centre and mycelium in the 
velamen. X 18 









"'■^c&jS'' 












ItiV- 







4^ 



• p., , 



'""■ ■ ■A.'i- 
lO 










.*;< 
'!(-ii 






12 








1'' _ 












.: J-*5 ^^ 






^^i:: •■■■■■ -'ftr 

15 






' M '■ Y.