12 Karl Hermann Fernow 



tomato still later, and potato only after a considerable period. When 

 inoculated with mosaic B, Datura Stramonium, Solatium aculeatissimum, 

 Nicandra Physalodes, and Nicotiana glutinosa, show symptoms readily, 

 while tomato and potato are slower. Datura Stramonium seems to show 

 mosaic B more quickly than mosaic C, and Nicotiana glutinosa appears 

 to show 'mosaic B only slightly more quickly than mosaic D. Mosaic H 

 apparently develops very slowly on both Nicandra Physalodes and 

 Nicotiana glutinosa. However, another investigator with a different 

 point of approach might recognize earlier symptoms on the suscepts here 

 mentioned as slow, or might overlook early symptoms on suscepts here 

 mentioned as showing symptoms early. 



SOURCES OF ERROR 



Any series of experiments such as is here reported is subject to certain 

 sources of error. The errors may be classified as errors in diagnosis, errors 

 due to accidental contamination, and errors due to the application of 

 methods which are ineffective for the particular disease and species of 

 plants under consideration. 



Errors of diagnosis 



Errors of diagnosis may be in either direction. There is abundant 

 evidence in literature, as well as in the writer's experience, that plants 

 may, under some conditions, harbor mosaic for long periods without 

 showing any visible symptoms. Such cases are reported for eggplant by 

 Melhus (1922), and for Physalis Alkekengi L. by Nishimura (1918). 

 If the plants are kept rapidly growing under favorable conditions, there 

 is less likelihood of error from this cause. The feeling is general among 

 plant pathologists that negative results are of little significance as com- 

 pared with positive results. While this feeling is justified in general, it 

 must not be forgotten that positive results may also be misleading. Posi- 

 tive results may be reported in the case of plants not actually affected, 

 due to the effects of low or high temperatures, insufficient water, infesta- 

 tion with thrips or red spider, or other conditions, or they may be due to 

 natural infection. 



In addition to positive diagnosis of plants affected with mosaic, some 

 attempt should always be made to identify the mosaic concerned. It is 

 evident that very few authors have made any attempt to determine 

 whether all the plants of a given species which they succeeded in artificially 

 infecting were attacked by the same virus. This would necessitate the 

 inoculation of parallel series of plants and a careful comparison of the 

 infected plants. Even when this is clone, the chances of an error in the 

 identification of the mosaic are greater than the chances of an error in 

 determining whether the plant in question is affected with mosaic. 



