STUDIES ON CLUHKOOT OF CRUCIFEROUS PLANTS 441 



For some unknown reason the amoeba in some cases may not penetrate 

 as far as the endodermis, but, after having reached a certain row of the 

 cortical cells, it may pass upward or downward in that row, cell division 

 taking place as fast as invasion occurs. This produces such rows of cells 

 as are illustrated in figure 100. The vertical direction may be changed at 

 any time and at frequent intervals to a horizontal one, and the adjacent 

 rows thus affected at once begin cell division in each direction. The result 

 is a true (< Krankheitsherde." Any one of these diseased rows, or several 

 of them, may extend beyond another group of healthy cells, from which 

 the organism again moves horizontally. This will cause a second " Krank- 

 heitsherde," above or below the first and separated by the length of one 

 or more healthy cells. This is illustrated in figure 106 better than it can 

 be described. A single infection may in this manner give rise to from 

 one to probably six "Krankheitsherde," with the intervening uninvaded 

 cells much increased in number over those in the normal tissue. 



It has already been pointed out why this longitudinal movement is 

 interpreted as the result of cell- wall penetration instead of as being due 

 to mere division. The diseased area shown in figure 106 is only five rows 

 of cells in width. The perfectly normal tissue of the same root shows 

 exactly the same number of rows. There has been no periclinal division, 

 and therefore direct migration must have taken place. 



All this occurs when the root is only a few millimeters in diameter. 

 For some reason the walls finally become impenetrable, and the amoebae 

 become more nearly globose and later are transformed into masses of 

 spores. 



SPORE FORMATION AND SIZE 



For the purpose of this discussion the nuclear phenomena need not 

 be included. The generally accepted explanation of a true mitotic division 

 followed by vacuolar separation into individual uninucleate spores is well 

 known. This separation is supposed to take place almost simultaneously, 

 but stained sections do not always show this to be true. Stages from the 

 amoeba to the mature spore are represented in figure 102, D (page 434). 

 In this case there were repeated successive separations instead of simul- 

 taneous fission, so that each amoeba is divided into two, theji four, and so 

 on until no two nuclei longer remain together. The unstained spores in 

 figure 1 02, A, show the same method of formation. They may then be 

 hexagonal or irregular in outline, and much larger than when mature, 

 but they soon become spherical. 



It was surprising to note the wide difference between the actual size 

 of the spores, and the measurements (i.6/z) given by Woronin (1878) 

 and nearly all succeeding authors. Molliard (1909) gives the diameter 



