60 
ducers generally little is known as yet. As a matter of course 
the winged species can fly from plant to plant, yet is it diffi- 
cult to assume that the extremely frail and slightly built gall- 
gnats should have flown from the. shores of the larger islands 
to Krakatau by their own unaided strength. One might of 
course imagine that the animals have drifted to the islands in- 
side their galls. This might be possible for instance in the case 
of the flower-gall on Derris uliginosa Benth., which grows in 
coastmarshes, and also of Clerodendron imerme, occurring in the 
same situation. But it is hard to see how this could happen 
where plants are concerned like Fices ribes and Nauclea pur- 
purascens, occurring as they do exclusively in the primeval 
forests on the mountains. How the animals on them have 
found their way from the virgin forests of Java and Sumatra 
to the islands, remains as mysterious to me as ever. But as 
nothing positive is known even about the manners and circum- 
stances of their spreading on the continents, a discussion of the 
possibilities is now premature, if even possible. We can only 
state that the occurrence of gnatgalls on Krakatau has proved 
that these small animals can spread even over great inter- 
vening distances. 
Concerning the spread of gall-mites better information is al- 
ready available. On this subject there appeared a few years 
_ ago an article whose author Natepa') is the specialist in the 
study of gall-mites. He argues that the spreading by their own 
activity is only possible over small distances, as they can move 
but slowly. A spreading to the islands by their own unaided 
efforts is of course out of question. 
But Narxpra also discusses two other means of spreading, in 
which the animals play a passive part: the gall-mites may 
Jags be carried by other insects or by winds. The former means ot 
eading was already mentioned by Warsurton and Emsueron *) 
liscussion of the enema of black-currants. In 
Ly A res oO acndie neuer Wirtsplanzen durch die Gallmilben, Mar- 
ia, Vol. 1X, 1910, 1 
f ARBURTON | and erage Lite ot cornet A of the — Carrent eats. Linn. 
urn. Zool. 4902, Vol. XX 
