468 BACTERIAL DISEASES OF PLANTS 



of gold chlorid. They are then washed 3 minutes in water and 

 placed for another 24 hours (in the dark) in a 0.25 per cent 

 aqueous solution of formic acid. After this they are washed in 

 water, passed through graded alcohols into xylol and embedded 

 in paraffin in the usual way. Is there not at least a strong prob- 

 ability that these rod-shaped bodies stained by the gold chlorid 

 and supposed to be the bacteria are only normal constituents of 

 the plant cell (mitochondria)? What are mitochondria? 

 (Read a paper by the Lewises in Journal of Anatomy, Vol. 17, 

 1915, p. 339.) Have you observed any bacteria in the inter- 

 cellular spaces of sound galls? Study the fauna and flora of 

 old galls. Can you find Tourney's organism? 



Have you observed any excess of chloroplasts in the tumor 

 or in the tumor-strand (daisy)? Any bleaching of tumors or 

 shoots from tumors ? Any floral pigment in tumors ? Try Pel- 

 argoniums, inoculating the tops of growing plants which are 

 nearly ready to develop red blossom buds. Try also red bal- 

 sams, inoculating before the flower buds develop. Any starch? 

 Any excess of sugar or of enzymes? Consult Figs. 353 to 356 

 for structure of the hyperplasial tumor tissue. Fig. 353 shows 

 spindle-celled tumor tissue and Fig. 354 shows round-celled 

 tumor tissue from the same gall. Fig. 355 shows both the 

 crushing and invasive effect of a tumor which is excessively 

 vascular, because arising from a very vascular organ the torus 

 of the sunflower. In Fig. 356 which is from Ricinus the glandu- 

 lar epidermis is also involved. When a tumor is deep seated 

 should the pushed-up and thickened cortex, the cells of which 

 are normally oriented, be reckoned as a part of the tumor? 

 If so, why any more than pushed up skin and muscle? 



Variability. We have found in various isolations from crown- 

 gall of the Paris daisy marked differences in virulence (ability 

 to produce galls), and from certain natural tumors on the sugar- 

 beet (supposed to be crown gall) none of the many typical 

 looking colonies on the agar-poured plates were infectious (we 

 tried perhaps a hundred). From other similar looking natural 

 beet tumors we obtained a very few infectious colonies, but 

 these produced only slow-growing small tumors (Bull. 213, 



