i^r'gi»'y^:?r^:/'"'Kif;"' SB 741 C6 C5 I Copy 1 'udies on clubroot of Cruciferous plants A THESIS PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF CORNELL UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY CHARLES CHUPP Published as N. Y. (Cornell) Agr. Exp. Sta. Bui. 387, 1917. STUDIES ON CLUBROOT OF CRUCIFEROUS PLANTS A THE5I5 PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF CORNELL UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY CHARLES CHUPP Published as N. Y. (Cornell) Agr. Exp. Sta. Bui. 387, 1917- 3>^<,^.W*<^.,>^ MARCH, 1917 BULLETIN 387 CORNELL UNIVERSITY AGRICULTURAL EXPERIMENT STATION STUDIES ON CLUBROOT OF CRUCIFEROUS PLANTS CHARLES CHUPP ITHACA, NEW YORK PUBLISHED BY THE UNIVERSITY CORNELL UNIVERSITY AGRICULTURAL EXPERIMENT STATION Experimenting Staff ALBERT R. MANN, B.S.A., A.M., Acting Director. HENRY H. WING, M.S. in Agr., Animal Husbandry. T. LYTTLETON LYON, Ph.D., Soil Technology. JOHN L. STONE, B.Agr., Farm Practice. JAMES E. RICE, B.S.A., Poultry Husbandry. GEORGE W. CAVANAUGH, B.S., Agricultural Chemistry. HERBERT H. WHETZEL, M.A., Plant Pathology. ELMER O. PIPPIN, B.S.A., Soil Technology. G. F. WARREN, Ph.D., Farm Management. WILLIAM A. STOCKING, Jr., M.S.A., Dairy Industry. WILFORD M. WILSON, M.D., Meteorolog>'. RALPH S. HOSMER, B.A.S., M.F., Forestry. JAMES G. NEEDHAM, Ph.D., Entomology and Limnology. ROLLINS A. EMERSON, D.Sc, Plant Breeding. HARRY H. LOVE, Ph.D., Plant Breeding. ARTHUR W. GILBERT, Ph.D., Plant Breeding. DONALD REDDICK, Ph.D., Plant Pathology. EDWARD G. MONTGOMERY, M.A., Farm Crops. WILLIAM A. RILEY, Ph.D., Entomology. MERRITT W. HARPER, M.S., Animal Husbandry. JAMES A. BIZZELL, Ph.D., Soil Technology. GLENN W. HERRICK, B.S.A., Economic Entomology. HOWARD W. RILEY, M.E., Farm Mechanics. CYRUS R. CROSBY, A.B., Entomology. HAROLD E. ROSS, M.S.A., Dairy Industry. KARL McK. WIEGAND, Ph.D., Botany. EDWARD A. WHITE, B.S., Floriculture. WILLIAM H. CHANDLER, Ph.D., Pomology. ELMER S. SAVAGE, M.S.A., Ph.D., Animal Husbandry. LEWIS KNUDSON, Ph.D., Plant Physiology. KENNETH C. LIVERMORE, Ph.D., Farm Management. ALVIN C. BEAL, Ph.D., Floriculture. MORTIER F. BARRUS, Ph.D., Plant Pathology. CLYDE H. MYERS, M.S., Ph.D., Plant Breeding. GEORGE W. TAILBY, Jr., B.S.A., Superintendent of Livestock. EDWARD S. GUTHRIE, M.S. in Agr., Ph.D., Dairy Industry. JAMES C. BRADLEY, Ph.D., Entomology. PAUL WORK, B.S., A.B., Vegetable Gardening. JOHN BENTLEY, Jr., B.S., M.F., Forestry. EARL W. BENJAMIN, Ph.D., Poultry Husbandry. EMMONS W. LELAND, B.S.A., Soil Technology. CHARLES T. GREGORY, Ph.D., Plant Pathology. WALTER W. FISK, M.S. in Agr., Dairy Industry. ARTHUR L. THOMPSON, Ph.D., Farm Management. ROBERT MATHESON, Ph.D., Entomology. MORTIMER D. LEONARD, B.S., Entomology. FRANK E. RICE, Ph.D., Agricultural Chemistry. VERN B. STEWART, Ph.D., Plant Pathology. IVAN C. JAGGER, M.S. in Agr., Plant Patholog>' (In cooperation with Rochester University). WILLIAM I. MYERS, B.S., Farm Management. LEW E. HARVEY, B.S., Farm Management. LEONARD A. MAYNARD, A.B., Ph.D., Animal Husbandry. LOUIS M. MASSEY, A.B.-, Ph.D., Plant Pathology. BRISTOW ADAMS, B.A., Editor. LELA G. GROSS, Assistant Editor. The regular bulletins of the Station are sent free on request to residents of New York State. CONTENTS PAGE Dissemination 421 vSpore germination 42,^ Penetration 427 Distribution within the host tissues . 434 Spore formation and size 441 A similar organism 442 Bacteria in relation to Plasmodiophora Brassicae 443 Summary 448 Bibliography 451 419 Pig. 95, DISEASED CABBAGE PLANT SHOWING THE THIN STALK AND THE ABSENCE OF A HEAD STUDIES ON CLUBROOT OF CRUCIFEROUS PLANTS^ Charles Chupp Such an extensive literature on clubroot of cruciferous plants has accu- mulated that it would seem impossible for any one point to have escaped careful consideration. - But when a close examination* is made of all the data, it soon becomes apparent that only such prominent phases as symptoms, cytology of the organism, and control methods, have been dealt with extensively, while certain other less conspicuous features have been neglected. There still remain to be satisfactorily solved the follow- ing problems: (a) the part played by swarm-spores in the dissemination of Plasmodiophora Brassicae Wor., the organism that causes clubroot; (b) spore germination; (c) the manner in which the pathogene enters the host; (d) the distribution of the organism thruout the tissues of the root; (e) formation and size of the spores; and (f) the relation of bacteria to the normal development of the myxomycete. It is for the solution of these problems that the following investigations have been conducted. DISSEMINATION In a general way the manner in which the spores are carried is known, altho two errors are often met with in popular descriptions. For example, in a number of reports (Atkinson, 1889, Carruthers, 1893, ^^^ others) ^ are statements implying that swarm-spores swim about in the water of the soil until they reach a cabbage root. In a way this is correct, but the average layman at once pictures the swarm-spores as traveling from row to row of plants or even from field to field. Nothing could be more erroneous, for, as far as dissemination is concerned, the motility of the swarm-spore plays such a slight part that it need not be considered. Its energy is not directed in a straight line, and the very minuteness of the organism woidd preclude any effective locomotion in the time that it remains alive. In order to test the distance to which swarm-spores may travel in the soil, a box two feet square was filled with clay mixed with muck soil, and diseased roots were buried in one end. Cabbage seeds were then sown in the box, care being taken not to transfer any of the soil from the place where the inoculum was inserted. When the seedlings over the area ' Also presented to the Faculty of the Graduate School of Cornell University, September, 1916, as a major thesis in partial fulfillment of the requirements for the degree of doctor of philosophy. Acknowledgment. The author gratefully acknowledges the helpful suggestions and criticisms offered him by Professor H. H. Whetzel and others in the Department of Plant Pathology at Cornell University. - Dates in parenthesis refer to bibliography, page 45 1. 421 422 Bulletin 387 Fig. 96. DISEASED CABBAGE SEEDLINGS Studies on Clubroot of Cruciferous Plants 423 where diseased roots were buried had become so badly infected that they began to wilt and turn yellow, all the plants were discarded and the plat was reseeded. Different crops of seedlings were thus grown for almost a year, and, altho there was a gradual spread of the organism, it was only by careless watering and planting that the pathogene was carried in the soil to all parts of the box. Cabbage seeds were sown in a greenhouse plat in rows ten inches apart, the bottom of each trench being first lined with infested soil. Halfway between these rows were sown other rows, in the, trenches of which no in- fested soil was placed. The inoculated plants (fig. 96) became infected at a very early stage, while the plants that were only five inches away from the spores remained healthy until they were almost mature. A few authors (Carruthers, 1893, and others) claim that wind is an important agent in spore dissemination. This may be true in light, loose soil, and in localities where strong winds prevail, but in none of the observations made by the writer was there a single case in which the presence of the organism could be explained on this basis. On the other hand, many of the fields showed that if the soil were not transferred by some agent other than the wind the pathogene did not spread. On Long Island, New York, a certain field was observed, one corner of which was slightly lower than the adjoining part. This corner had been used for a garden until clubroot became so prevalent that the plat was no longer profitable for the raising of crucifers. It was then tilled with the remainder of the field for three years while various crops were grown, cabbages not being planted again until the fourth year. A space only slightly larger than the original garden then displayed the presence of clubroot. If wind had been an important agent it would have had an opportunity here, for the land was almost level and the soil was very loose. This was only one of several, cases in which the same conditions were observed. SPORE GERMINATION Very few persons have been successful in germinating the spores of Plasmodiophora Brassicae, and of those few who have been so fortunate, still fewer have seen the actual process. Woronin (1878) gives a brief description and a series of illustrations which have been copied by nearly all later writers on this phase of the subject. The general experience, however, seems to have been like that of Maire and Tison (19 11) while working with Tetramyxa parasitica Goebel. They saw only one spore actually germinating, and after a very long, tiresome vigil they left it for a few minutes. On returning from their temporary absence they found that the phenomenon had been completed. Notwithstanding these diffi- 424 Bulletin 387 culties, Eycleshymer (1894) not only found swarm-spores, but also found that when left in the culture for a few days these apparently fused into larger bodies, thereby reacting in much the same manner as Kunkel (191 5) found to be the case with Spongospora subterranea (Wollr.) Johnson. Kunkel discovered that each cell of a spore ball produces a single uninucleate amoeba which soon fuses with others of its kind to form a small plasmodium. This occurs not only in the case of spores in the soil, but even with those still in the base of the old sorus. There are several obstacles to be encountered in trying to observe the actual emergence of the protoplasmic mass from the old spore wall. First, it is difficult to get a very large percentage of germination unless the most favorable conditions are present. Secondly, all observations must be made with the oil-immersion objective. When the protoplasm is about half- way out, the spore wall and the emerging protoplast begin to move, making it hard to keep them in focus or even within the field. Consequently, when the process seems almost complete there is a sudden swift whirl, and the swarm-spore, with the adhering empty wall, darts out of sight. When located again, the spore wall is empty, and the swarm-spore, lost among others, is impossible of identification. For this reason no actual separation of the protoplasm from the spore wall has been seen, but enough of the process has been observed to enable investigators to deter- mine the general method by which this is accomplished and to be sure that a spore gives rise to only one swarm-spore. It was soon learned that spores do not germinate well, if at aU, in dis- tilled water, and further that, altho from one to five per cent of the spores taken directly from a fresh root germinate in muck-soil filtrate, a much larger percentage of germination can be obtained by exposing the roots to freezing temperatures for two weeks or longer. This was accomplished by tying the roots in cheesecloth and burying them under the snow, or in summer by keeping them in the refrigerator for that length of time. Drying the roots also seems to have a beneficial effect on germination, altho this must not be carried to the extreme. The muck-soil filtrate was made by filling an ordinary flowerpot with muck, placing it over a large funnel lined with filter paper, and then pouring hot water on the soil. The resulting medium was of an amber color and slightly acid. Temperature conditions also influence germination of the spores. It was practically impossible to obtain infection in the greenhouse during the coldest winter months when the temperature was from 10° to 18° C. The spores also fail to germinate at ordinary room temperature (from 1 6° to 21° C). The optimum temperature for germination proved to be from 27° to 30° C. This, however, is not the case when spores are placed in test ttibes on agar with young cabbage seedlings, for under such conditions Studies on Clubroot of Cruciferous Plants 425 infection takes place at a temperature of from 16° to 21° C. The pres- ence of the host seems in some manner to exert an influence which to a certain extent takes the place of that offered by a greater amount of heat. Usually the first sign of germination is a swelling of the spore, which sometimes becomes a third larger. This occurs within a period of from fifteen minutes to eight hours after the spores are placed in the medium, altho the best time for examining the culture proved to be at the end of six hours. After the swelling of the spore there is a bulging at one side. The protoplasm withdraws from near the opposite wall and leaves a nearly hyaline semicircle about two-thirds of the distance from the center. The pressure exerted splits the wall just enough to permit the protoplasm to ooze out. Unlike Woronin (1878) and Mangin (1902), the writer has never observed the protoplasm taking the various shapes that these authors assign to it, but while oozing out it collects in a sphere or a hemi- sphere against the wall on the outside. When about half of the proto- plasm has escaped, the whole body becomes motile. At first there is only a trembling, which gradually increases in violence until the spore is turned around entirely. The activity now becomes so great that it is with diffi- culty that the microscope is kept focused on it correctly. The final struggle is apparently a rapid spurt across the field, when the swarm-spore is lib- erated from its container and at once begins its rotatory activities. The whole process under the microscope consumes an hour or longer. Evi- dently the strong light turned on a spore retards the action, for in many cases the spores that had begun to germinate when placed in view showed no further signs of development, while those kept in the dark germinated much more rapidly and when examined at the end of the same period were found actively swimming about. A considerable part of the contents is left within the old spore wall, so that when the broken part is turned upward it has the appearance of a circle bounded by a darker band, the width of which is about one- third of the radius. If, however, the open part is on the side, the residue within the spore wall resembles more nearly a crescent (fig. 97). The swarm-spore when alive measures from r.7 to 3.5 m in length, being more or less pyriform with a thick flagellum at the smaller, or anterior, end and a vacuole near the posterior end. Unless stained, the flagellum cannot be seen under the microscope. The line of locomotion is never a straight one, for the flagelltmi is lashed about by the beak, which is constantly doubling backward so that a whirling motion is given to the swarm-spore. Altho the latter is a naked mass of protoplasm, the writer has never seen the various shapes which Woronin (1878, PI. xxxiv) has pictured; it was observed in every case to be globose or pyri- form, never having pseudopodia-like structures. 426 Bl'LLETIN 387 It has been difficult to properly fix swarm-spores for staining flagella. The first method of staining tried was that ordinarily employed for bac- teria, namely, Loefflsr's mordant and Ziehl's carbol fuchsin. When bacteria were in the mount their flagella were stained, but those of the «P5^ swarm-spores had evi- m dently disappeared. The process was then modi- fied slightly, and the cover-glass mounts, in- stead of being left to dry in the incubator, were placed on slides in preparation dishes with ground-glass tops. In the bottom of each dish was placed a few cubic centi- meters of osmic acid, and the lid was then carefully fitted in place. The acid killed a few of the swarm- spores before the flagella could be withdrawn, but never a very large pro- portion. Besides demon- strating the presence of flagella, the stained ma- terial also displayed different stages of ger- mination (fig. 97). Kunkel ( 1 9 1 5) was able to get spore germination of Spongospora subter- ranea on an agar me- dium. Plasmodiophora Brassicae evidently does not react in the same way. During the three years of the present work, repeated efforts were made to secure not only germination on the surface of agar, but also formation of plasmodia. Unless the spores were immersed in water there was no development. They lay there until the agar became so dry that they finally lost their viability. If enough of the muck-soil filtrate was added, the swarm- Fig. 97. SPORES and swarm-spores of plasmodi- ophora BRASSICAE The two spores at the top have already germinated. The germi- nating spore and the two swarm-spores near the bottom were drawn from stained mounts. The bacillus shown is the form found oftenest in older diseased roots. X 2100 Studies ox Clubroot of Cruciferous Plants 427 spores appeared but there was no further development. They were active for a certain time, and then encysted and remained in that con- dition as long as the cultures were kept. This experiment was performed on four kinds of agar media, on potato plugs, and on healthy cabbage roots. In no case were there any signs of further growth. This, with subsequent infection experi- ments, indicates very strongly, if it does not prove positively, that the swarm-spores never fuse. This is in keeping with what has been found, or at least suggested, in all other cases of parasitic ^' slime molds, Spongospora suhterranea excepted. ] If spores for germination are taken from roots „ ,, , , , . , . -,■ ■ r ■, -, Fig. 98. FLAGELLATE OR- that have not previously been dismrected, there ganisms a.ssociated are often found in the cultures flagellate bodies ^^'^th plasmodiophor.^ ... , ,, , , , BRASSICAE which are almost small enough to resemble swarmspores. They are larger, however, are more active, and when stained are more or less reniform, having two fiagella arising from the concave side (fig. 98). These, as pointed out later, belong to another organism. PENETRATION In the knowledge of the life history of Plasniodiophora Brassicae, there has always been a gap between the swarm-spore stage and the amoeba within the cell, the true sequence of development never having been shown. Most writers pass over the difficulty with the mere statement that the organism enters the root and there begins its parasitic life. Woronin (1878), in this as in nearly all other points connected with clubroot, is the only one who has tried to fill in the gap. In a way he succeeded, but, as his plants died before reaching the stage in which invasion of any of the tissue took place, he is not sure that the root hair is the real point of entrance. He placed cabbage seedlings in shallow watch glasses, in water well supplied with spores. For some reason the plants began dying before hypertrophy took place. When the roots were examined microscopically, the root hairs were filled with amoebae but nothing further had happened. The question still remained, whether these infections under normal con- ditions would have been followed later by invasion of the cortical cells, or whether the case was similar to that which Schwartz (19 14) found in species of Ligniera. Schwartz thinks that penetration takes place near the apex of the root, so that when the root hairs act as bearers of the amoebae the parasite does not advance farther than the base of the cell. Most writers believe not only that the apical cells and the root hairs act as infection courts, but also that the epidermal cells can be infected 428 Bulletin 387 directly up to the time when the epidermal layer is thrown off (Woronin, 1878). Somerville (1895) gives an observation as proof of this statement. He often found swellings high up on the roots of turnips, where he declares no root hairs could have been responsible for the entrance of the slime mold, which must have penetrated the thick cuticle. This question of entrance has a direct economic bearing on control, for, if Somerville's statement is true, Massee's (1903) assumption is certainly erroneous. Massee states that the Cruciferae can be attacked only during seedling time, and that after six weeks they are practically immune. It is doubtftd whether either Somerville or Massee interprets the conditions correctly. If infection could not take place after six weeks, the grower could control the disease merely by late transplanting and the proper care of his seed beds; but this has evidently not proved to be the case in practice. Maire and Tison (1909, 191 1) and vSchwartz (1910, 191 1, 1914) have done nearly all the work that has been reported on the parasitic slime molds other than Spongospora subterranea and Plasmodiophora Brassicae. It is interesting to note that their conclusions agree very closely, and that they feel sure the amoebae enter oftener thru the apical cells than otherwise, altho the root hairs also may serve as points of entrance. They made no particular study of this question, but were led to this conclusion by finding uninucleate amoebae in the cells near the growing tips. Their opinion is substantiated also by the presence of rows of diseased cortical cells, the divisions of which apparently take place when still very near the initial cells in the root tips. The powdery scab pathogene, Spongospora sub- terranea, passes directly thru and between the epidermal cells into the tuber (Kunkel, 19 15). There is more or less difhctdty in studying the nature of penetration in the case of Plasmodiophora Brassicae, because of the fact that the uninucleate amoebae are so small. They can be recognized only under a very high magnification, and, since they are so nearly transparent, stained sections must be used for all the work. A very large number of both longitudinal and cross sections were prepared, the thickness ranging from three to fifteen microns, and the staining was done with the combination stains of safranin, gentian violet, and orange G. These proved best for differentiating the parasite from the host, especially when orange G was used in excess. There is no possible stage in penetration that was not represented in the preparations. Large, as well as very small, roots were sectioned, and a great number of epidermal cells showed amoebae. But in a careful study of almost three hundred slides, none of these cells showed that penetration had taken place directly thru the cutinized wall. In a number of cases this appeared to be true when the sections were first examined, Studies on Clubroot of Cruciferous Plants 429 but a more detailed study of the same series showed the invaded cell to be in every case the basal portion of a root hair. This, together with the fact that no new swellings are ever found at any great distance from the region where root hairs might have existed previously, has led the writer to believe that seldom, if ever, is there direct penetration into simple epidermal cells. This holds true not only for the area above the place where the root hairs have disappeared, but evidently also for the space near the extreme 'tips where the hairs have not yet been formed. Not only did these slides demonstrate this point, but infection secured under aseptic conditions in test tubes has confirmed it. The small root-tips were so placed that they were the first to come into contact with particles of diseased tissue and the muck-soil filtrate containing free spores. When these rootlets were sectioned and stained, they showed various stages of root-hair invasion, but no amoebae were found in any of the apical cells. The evidence pre- sented in these slides shows that these invasions are not, like those which Schwartz (19 1 4) suggested for Ligniera sp., confined alone to the epidermal cells of which the hairs are outgrowths. The passage of amoebse from the epidermal cells into the cortical tissue is demonstrated not only by the position of the amoebae within the paren- chyma cells, but also by actual cell- wall penetration. The argument advanced for other species of Plasmodiophoraceao, that infection must take place in the growing tip where cells are dividing rapidly because the organism often occurs in definite rows of the cortical cells, does not necessarily apply to Plasmodiophora Brassicae. A glance at a section of a root tip (fig. 99) indicates the difficulty that a swarm- spore woiold encounter in entering at this point. The rootcap does not merely protect the root tip, but a row of its cells extends upward almost halfway to the root hairs. The remaining distance is protected by a comparatively heavy cuticle, leaving the root hair as practically the only vulnerable point. Moreover, the presence of the organism in continuous rows of cells can be explained in another manner. The condition shown Fig. 99. LONGITUDINAL SECTION OF A CABBAGE ROOT This shows the tip of the cabbage root protected by the cells of the root- cap. X 1 10 43° Bulletin 387 in figure 100, b, gives no indication as to where penetration occurred. Yet by moving the section the length of half a dozen cells, there is seen an uninterrupted connection of diseased tissue between this particular row and the epidermis (fig. 100, a). So far as the writer's observations go, there seems to be no question but that penetration does take place thru the root hairs, and thru these only. Eycle- shymer (1894) suggests that wounds caused by insects may provide a means of entrance for the parasite. This is altogether probable ; yet the writer has never observed any indica- tions of this condition, so that if it ever hap- pens it apparently does so very rarely. If cul- tures coiild be secured within pieces of healthy disinfected roots in test tubes, it would at least be evidence that such wound infection might take place. Pinoy(i9o5) removed small pieces of nealthy roots by means of sterilized pipettes, and by inoculating them with spores secured cul- tures of the organism. Fig. 100. DISEASED CORTICAL TISSUE OF A CABBAGE ROOT A, A row of diseased cortical cells; B, another row of diseased cor- tical cells connected with the epidermis by an unbroken hne of diseased orovidcd th C tubcS WCrC tissue. X no ^ sealed so that the aerobic bacteria were deprived of oxygen. His discussion of this point is some- what lacking in clearness. Besides, the time in which he claims spores were produced in the roots is unusually short. He gives it as five days, Studies on Clubroot of Cruciferous Plants 431 which is the same time that under the most favorable circumstances it takes swarm-spores to pass thru the root hairs into the cortical tissue and to develop sufficient hypertrophy to be visible to the naked eye. Kleimenov (19 12) tried the same experiment and failed. In the writer's experiments it was also tried repeatedly, always with failure. If the cul- tures were kept free from bacteria the root underwent no change. If bacteria were added, the root became soft and foul-smelling, whether the test tubes were closed with cotton plugs or sealed with paraffin over cork or cotton stoppers. Sealing did not stop the growth of the bacteria, as Pinoy claims for his experiments. Altho authors poptdarly describe with some assurance various ways in which the organism may enter the host, no one has observed the real process. Even Woronin, who believed that the organism passes thru the root hair, was never able to demonstrate this clearly. Nevertheless he felt assured that it enters in the form of a uninucleate amoeba, and his opinion has been accepted by most investigators. A few workers, such as Worthington G. Smith (1884), maintain that the organism enters the root in the form of a Plasmodium, but this theory has never been accepted generally. The question was revived again when Kunkel (1915) studied the powdery scab of potato, in which the swarm-spores are found to fuse before attacking the host. There seems to be no doubt in the minds of Maire and Tison (191 1) and Schwartz (19 14) that all the other known parasitic myxomycetes enter immediately after the swarm-spore stage. This conclusion is based on the fact that many of the slides of these investigators show the uninu- cleate forms in the apical cells. There is no other theory that would explain this phenomenon, unless a single uninucleate amoeba of an infecting Plasmodium passes thru the intervening cell walls and spreads in this manner thru the tissue. This is improbable. Because of the diminutive size of the swarm-spore, the only satisfactory method for studying penetration appears to be by means of stained sec- tions of roots showing the earliest stages possible. In the first part of this work, young plants from the greenhouse were used, but none of the stages were young enough to give just what was desired. An attempt was then made to grow plants in large test tubes on screens so arranged that the roots were hanging in muck-soil filtrate containing a heavy suspension of spores. The roots did not develop well when immersed in the liquid medium, and but few root hairs were present. An attempt was then made to grow seedlings in soil, in flats six inches square, with diseased tissue so plentiful that none of the plants could escape infection. The roots were fixed and embedded at intervals before the time when ordinary symptoms became apparent to the naked eye. This gave nearly all the early stages 432 BVLLETIN 387 of infection, but the adhering particles of soil, which eoiild not be washed off without sacrificing the hairs, not only were detrimental to the micro- tome knife, but also obstructed a clear view of the cell walls. Finally a method was devised whereby infected roots could be procured free from any other contamination. Diseased roots that contained spores but were not far enough advanced to be invaded by bacteria "were sterilized on the surface with mercuric chloride and transferred to agar slants in test tubes. After two weeks cooling in the ice chest they were finely minced in the agar, and incubated until it was clear that no bacteria were present in the tissue, from which they might have been liber- ated by the cut- ting. After enough time had elapsed to insure perfect freedom from any saprophytes, a few drops of sterilized muck-soil filtrate, and a young cab- bage seedling which had been grown from disin- fected seed on agar in a petri dish, were added. It was necessary to exercise care in adding sufficient liquid to permit spore germination and not have an excess, which would injure the root. A few drops would not evaporate until all the swarm- spores had ample time to be set free and attack the root hair. The process was somewhat long, and very often roots were chosen which were too old and were already contaminated with bacteria. In spite of all the difficulties, enough pure cultures were obtained to provide a large number of sections which showed all sizes of amoebse. The first and most important thing shown by the stained sections was that Plasmodiophora Brassicae enters the root hair as a uninucleate amoeba, Fig. 1 01. THE AMCEBA OF PLASMODIOPHORA BRASSICAE IN A ROOT HAIR A, A root hair with an amcEba showing two nuclei. B, A uninucleate amoeba in a root hair which shows an abnormal swelling in the immediate vicinity of the organism. C, A uninucleate amoeba in a tangential section of a root hair; the nucleolus has elongated, as it ordinarily does just before nuclear division. D, A host nucleus in a root hair, showing its size as com- pared with that of a uninucleate amoeba. E, A uninucleate amoeba in a shrunken, distorted root hair. X 1600 Studies on Clubroot of Cruciferous Plants 433 not as a Plasmodium. There are several facts that prove this conclu- sively, even tho the actual phase of the organism passing thru the wall was never observed with certainty. A number of slides show cases that might be interpreted as actual penetration, but as the nucleus in no case appears in the act of making the passage one cannot be certain of such an interpretation. Nevertheless, numerous cases are to be found of a uninucleate amoeba just within the wall of the root hair and far enough away from any other infection to preclude all possibility of its having reached there except by entering singly thru the wall (fig. loi). Evidently the reason why no one has recorded this stage heretofore is because the amoeba hardly enters before nuclear division and growth takes place. Some slides show binucleate amoebae still within the hollow of the enlarged cavity, apparently produced by the stimulus of the para- site. Other sections show trinucleate amoebae, and it is not difficult to find amoebas with six or more nuclei (fig. 104, page 436). This series of stages would indicate that penetration takes place in the uninucleate stage, particularly since the large multinucleate amoebas are to be found, in nearly every instance, near the base of the root hair, while the smaller and fewer-nucleate amoebae are always on the inside of the root-hair wall about two-thirds of the distance from the base. Amoebae are seldom found in the tip of the hair. Another point that confirms the above view of penetration is that in the absence of growing host roots the swarm-spores develop no further when the spores are germinated under artificial conditions, and after a short period of activity the swarm-spores encyst and eventually die. If Plasmodia are formed under normal conditions, there should have been at least a suggestion of this in a few of the numerous cultures used in the experiments. In this connection also the very interesting question of sexual fusion arises. It is believed by several cytologists that there are two nuclear divisions just before spore formation and that one of these is probably a reduction division. If this is true, it would imply that somewhere in the life cycle there has been a fusion. Winge. (1913) and others believe that this occurs among the swarm-spores before they enter the host. Prowazek (1905) is of the opinion that the amoebae within the host unite and then the nuclei fuse. Even Nawaschin (1899) believes this union takes place, but apparently he thinks it is of no significance in reproduction. Maire and Tison (1909, 191 1) have disproved the amoebal union, and their view is certainly correct, for it is possible to find slides showing one amoeba breaking up into spores while in another, immediately adjoining, di\nsion 434 Bulletin 387 has not yet begun (fig. 102, d). On the other hand, it would seem that the fusion of two swarm-spores would give an increase in size, but the measure- ments of amoeba just after penetration show them to be no larger than the swarm-spores just out of the spore wall. Consequently Winge's theory Pig. 102. SPORES and amceb^e of plasmodiophora brassicae A, Spores before their final separation from one another; B, cell filled with amoebae; C, cell filled with spores^ All X 800. D, Formation of spores, X Soo must be discarded. It thus appears that the real fusion stage, if there is one, is still to be discovered. distribution within the host tissues As stated above, the uninucleate amoeba, just after its entrance into the host, lies at first in a small cavity produced by the outward swelling of the part of the root hair at the point where the organism entered. This protuberance is no doubt caused by the irritating presence of the para- site (fig. 10 1, A, B, e). Following penetration the amoeba increases in size and pushes toward the center of the hair. The movement is accomplished by an actual amoeboid creeping, and an elongation and gradual segmen- tation of the forward ])art. W.oronin (1878) was able to observe the Studies on Clubroot of Cruciferous Plants 435 former method of locomotion in the living cells, and mentions it as the means by which the organism moves. Schwartz (19 lo), on the other hand, observed the growing of the amoeboid tip in Ligniera Junci (Sch.) M. et T., and explains the change of position on that basis alone. A root hair is shown in figure 104, d, which apparently was infected near the tip, and as the organism grew rootward fission took place, so that when the anterior part of the amoeba eventually reached the base of the cell the root hair was filled completely with the meronts, as Maire and Tison (191 1) designate the segmented parts (fig. 103). This does not always take place, for there were many more cases observed in which the intact amoeba reached the base of the cell (fig. 104, e, f). In either case, if the time consumed is too long, or if | — - — ^^^ for any other reason sponilation L^ ' j begins, the amoeba loses its power of further penetration into the cortical tissues. If, however, it reaches the inner wall of the root-hair cell, its pseudopodia are extended into the very smallest thread-like ])rocesses, which pass thru and into the cortical cell (fig. 105, E, f,g). Schwartz (19 10), in de- scribing penetration by Ligniera Junci, gives the same route of invasion but does not state how the passage from the epidermis into the cortical cells takes place. This question is of es- pecial interest, since in the latter part of his discussion Schwartz states his belief that amoebae never have the power of penetrating cell walls. There is no other apparent means by which this could be accomplished, for the epidermal cells seldom divide periclinally. It would be difficult to explain the wide distribution of the parasite within the root if cell-wall penetration did not occur, even tho it were taken for granted that invasion begins in the apical cells. The rootcap S3 fully protects these rapidly dividing primary cells that one must pre- suppose that in order to reach them the organism can pierce the walls. Then, in the maturer roots constant secondary thickening by the cam- bium takes place, which would ultimately push most of the diseased cells toward the periphery or isolate them near the center. This, however, does Fig. 103. PHOTOMICROGRAPH OF CELLS CONTAIN- ING AMCEB^ One amoeba has elongated considerably and is separating into meronts 436 Bulletin 387 '- ,