Dec. 23, igis PavasiHsm of Cronartium rihicola 623 



with 50 per cent hydrofluoric acid and then embedded in celloidin/ or 

 was simply placed in 95 per cent alcohol and glycerin, equal parts, for 

 about 10 days to 2 weeks, and sectioned without embedding at all. 

 Neither of these two methods gave better sections than could be obtained 

 from fresh tissue with the aid of the ether freezing microtome and a sharp 

 knife, except in the case of tangential sections. A thickness of 10 to 20 ,u 

 was found the most favorable for the study of the mycelium in the differ- 

 ent elements of the cortex, phloem, and xylem. In sections thinner than 

 10 /x the mycelial strands were often torn out of place. For the study of 

 the pycnia and aecia sections were used as thin as 3 n, but 5 to 7.5 ix v^'ere 

 generally employed. Sections of uredinia and telia were cut 3, 5, and 

 7 IX thick. 



For mounting. Land's fixative,^ potassium-bichromate + gum-arabic, 

 was found superior to egg albumen and was used almost entirely for bark 

 sections. It was also found more satisfactory for long sections of the leaf. 



A rather long series of stains was employed, and a comparison of the 

 different features of the mycelium and host cells made from slides colored 

 with the different stains. The diagnostic method has already been 

 mentioned. Alcoholic safranin and clove-oil -f gentian-violet were par- 

 ticularly good for mycelium in the xylem. Safranin and Delafield's 

 hematoxylin was also a favorable combination for such infected tissue. 

 In the phloem and cortex the mycelium was well differentiated with 

 Delafield's hematoxylin followed by erythrosin in 70 per cent alcohol. 

 For cytological study Haidenhain's iron-alum hematoxylin in combina- 

 tion with aqueous Congo red, aqueous orange G, or Lichtgriin in 95 per 

 cent alcohol proved excellent. Host and parasite cell walls stained well 

 with the Lichtgriin, whether in combination with the hematoxylin or 

 safranin. Flemming's triple stain ^ gave the best results for cytological 

 study, with one or two exceptions to be mentioned later; and by using a 

 strong violet stain the mycelial and host relations at the bases of the sori 

 were brought out more clearly than with any other combination. The 

 gentian-violet was made up in small quantities, enough for one week, as 

 it does not keep well. All reagents from the safranin to the xylol were 

 handled with pipettes from dropper bottles and used but once. 



It was found that material once correctly embedded in paraffin, sec- 

 tioned, and mounted, would stand the jump from 95 per cent alcohol to 

 water, then a 5-minute immersion in full strength hydrogen peroxid for 

 bleaching, wdthout shrinkage or distortion. The sections were rinsed in 



1 See Plowman, Amon B. the celloidin method with hard tissues. In Bot. Gaz., v. 37, no. 6, 

 p. 456-461. 1904. 



' See Chamberlain, C. J. methods in plant histology, ed. 3, 314 p., illus. Chicago, 1915. 



3 The stains were made up according to the followiiiK directions: Safranin, Solution I: 2 gm. of safranin 

 in 95 cc. of 95 per cent alcohol; filter. Solution II: 5 cc. of anilin oil in 95 cc. of distilled water. Shake the two 

 solutions together and filter. Guntian-violET, Solution I: i gm. of gentian-violet in 15 cc. 95 per cent 

 alcohol. Solution II; 3 cc. of anilin oil in So cc. of distilled water. Mix the two solutions, shake, and filter. 

 Orange G: i gm. of orange G. in 100 cc. of distilled water. 



