Ramaley: anatomy of onagraceae. 679 



origin of the islands. In the majority of cases, according to 

 this investigator, the intra-xylar islands have their origin in 

 the cambium, and are developed exclusively in a centrifugal 

 manner. Oenothera furnishes an example of this method of 

 formation. He states that there are but three genera known 

 at the present time, in which the phloem islands are developed 

 centripetally. These are Strychnos, Memecylon, Guiera. 



The latest contribution to the subject is by Perrott, 35 who 

 has made a careful study of the development of the phloem is- 

 lands in Strychnos. He confirms, in general, the observations 

 of Scott and Brebner, though disagreeing in some points. No 

 reference is made to any species of Onagraceae. 



Methods. — For most of the work carefully determined herb- 

 arium material was employed. Of the Minnesota species sec- 

 tions from fresh material were also examined. Whenever pos- 

 sible a number of specimens of each species from different 

 localities were used. Portions of the stem were taken at dif- 

 ferent heights. Some of these pieces were placed in 2 per 

 cent, potassium hydrate solution for twelve hours and, after 

 being thoroughly washed in water, were put through the usual 

 process preparatory to section cutting. Other pieces were 

 put at once into 10 per cent alcohol and from that carried up 

 through the higher grades of alcohol in the ordinary manner. 



Various stains were used. The most satisfactory ones tried 

 were a watery solution of fuchsin and methyl blue, and a solu- 

 tion of fuchsin and iodine green in 40 per cent, alcohol. Both 

 these combination stains give good differentiation. The best 

 results were obtained by employing the stains in very dilute 

 form, allowing them to act from twelve to twenty-four hours. 

 By using an eosin-haematoxlin stain some excellent prepara- 

 tions were also made. Permanent mounts in Canada balsam 

 were preserved of all the sections examined. 



In order to dissociate the separate elements for careful study 

 pieces of stem previously soaked in water were treated with 

 Schulze's macerat'mg mixture. When diluted with an equal 

 quantity of water this mixture was found to be sufficiently 

 strong. Only in macerated material could the length of stere- 

 ids and wood fibres be at all accurately determined. 



For examination of the epidermis it was merely necessary to 

 place thin portions of the cortex in water for from one to two 

 hours, when the desired tissue could be peeled off with little 



35 Perrot. Sur les ilots libcriens intraligneux des Strychnos.— Morot, Journ. de 

 Bot.9:90. 1896. 



