ON THE EFFECTS OF URBAN FOG UPON CULTIVATED PLANTS. 19 



that some poison is at work. Nor is it possible to say that any par- 

 ticular part exhibits this earlier than the rest. The cuticle is not 

 effective here in retarding its action from outside, and it is con- 

 ceivable that the epidermis absorbs the poison directly, or this 

 may be supplemented by its entry into the intercellular spaces. 

 Nor are leaves which fall green, or those which change slowly but 

 uniformly in colour, of much service in determining the point at 

 issue. But in leaves of Bouvardia, Carpentaria, some Bhodo- 

 dendrons, and a few others in which the process has been 

 carefully followed, I find that it is the cells of the spongy 

 parenchyma which are first affected. It is in the lower layers 

 of this tissue — in the parts adjacent to the stomata — that the 

 action commences. It then spreads to the upper regions of the 

 spongy parenchyma and to the palisade layer, and to the 

 epidermis of either surface. When the epidermis is cuticularised 

 it takes longer for the noxious vapour to traverse the cuticle than 

 it does for it to enter by the stomata and reach it this way. But 

 where the epidermis is very soft both methods may prevail. I 

 had come to the general conclusions indicated in this paragraph 

 from the examination of fog-injured foliage only, and before ever 

 I had experimented upon the action of various vapours. It will 

 be seen from what appears below that these conclusions receive 

 confirmation from the observed mode of action of a variety of 

 substances. 



Analyses of Injured and Uninjured Leaves. 



The very greatest care and deliberation is necessary before 

 one can base conclusions on the results of the chemical analysis 

 of foliage, &c. ; consequently I have decided, for the present, to 

 keep back the many determinations of this kind which — with 

 Dr. Bailey's co-operation — I have been able to obtain.* 



Special paragraphs dealing with changes in the chlorophyll 

 and with the injuries exhibited by flowers are deferred until the 

 action of various vapours has been described. 



* Cf. Just u. Heine in Landwirthsch. Versuchsstat. 1889, for a dis- 

 cussion of the difficulties attending this method of investigation. 



c 2 



