Manchester Memoirs, Vol. liii. {igog), No. *Z^. 3 



the midrib, and this results in the curling of the leaves 

 ventrally, the upper side of the leaf being outwards, as 

 seen at br., where the young leaves have begun to be 

 inrolled. On the left-hand side of the figure a twig is 

 seen, which has not yet been enclosed by a detritus casing. 

 Here the leaves are more tightly rolled up as they grow 

 older, and the space formed by the inrolled portion is 

 enclosed with sand. The ants are found in these inrolled 

 leaves. When fresh, the leaves appeared to be living, as 

 no doubt was the case, otherwise the aphides, for whose 

 protection the tents are constructed, would be unable to 

 subsist on them. They had, however, no trace of green 

 colour, but were instead a livid purple. The leaves which 

 project from the tent, as //., had green tips. A normal 

 untouched leaf is shown at 71. 



Unfortunately, as the observations were made during 

 a few days vacation, no reagents or instruments were 

 available, and no field experiments could be made. The 

 specimens collected were fixed in the local wine sake', 

 which proved to be an excellent preservative, as its per- 

 centage of alcohol is very high. 



In their anatomy the dwarfed and rolled leaves show 

 no fundamental difference from the normal ones, but all 

 the tissues, particularly the vascular tissue and palisade 

 cells, are much less differentiated. For example, one 

 layer only of short palisade cells in them represents the 

 elongated double layer that is found in the normal leaf. 

 Along the under side of the midrib are several smaller 

 wounds, which remain as brown scars of destroyed tissue, 

 as seen under the microscope ; these are probably pro- 

 duced by the biting action of the ants' mandibles, and 

 cause the rolling of the leaves. Brown scars also occur 

 along the edge of the leaf, which in some instances is 

 slightly toothed. 



