2G 



The pièce without tlie midrib was examined at once, 

 that containing the midrib was used for the experiment 

 It remained on the tree, was put in water, laid on moist 

 blotting paper or treated in other ways. At the end of the 

 experiment the midrib was eut off and the remaining half 

 of the leaf examined. 



In this way the leaf- parts to the right and the left of 

 the same midrib were always compared with each other. 



For examination the leaf- parts were eut into very small 

 quadrates and boiled in alcohol containing 1 / 2 % of HC1 (20 

 ce. conc. HC1 on the Liter) for an hour. Thistook place on 

 the watérbath in Erlenmeyer-bottles, stoppered by a co'rk 

 into whicha long glasstube, serving as a cooler, was put. The 

 alcohol was afterwards poured into porcelain dishes, placed 

 on the watérbath and the whole evaporated until nearly 

 dry. Afterwards the dishes were filled up with water and 

 evaporated aga in until nearly dry also, so as to be sure of the 

 total escape of 'the alcohol. After this water was added, again 

 filtered and the filtrate collected in a separotory. After 

 addingeaustic potash until alkaline solution it isshaken with 

 chloroform, the chloroform collected in a watchglass, put on 

 the watérbath and ail chlorofrom evaporated. The résidu 

 is dissolved in water, containing 1 / 2 °/ Q of HC1 (20 ce. conc. 

 HC1 to the Lifcer.) By strong rubbing the remious 

 substances, sticking to the watchglass are mixed with this 

 solution, the whole filtered and the filtrats used for the 

 alcaloid tests. I followed the chemical part of this method 

 owing to the advice, kindly given by Dr. W. G. Bjorsma to 

 whom my sincère thanks are here offered. 



At the commencement of this investigation nearly ail 

 the usual alcaloid reagents w T ere used. When alcaloid was, 

 présent they ail gave sumptuous précipitâtes. To décide 

 wether the leaves were empty or not thèse reagents fre- 

 quently were too sensitive, even the very smallest quan- 



