360 Mr F. Darwin and Mr R. W. Phillips, [Nov. 23, 



IV. The experiments with coloured fluids tend to confirm our 

 belief in the similarity between the transpiration and filtration 

 streams. 



We have hitherto considered our results chiefly with reference 

 to Dufour's experiments. We will now discuss them in a more 

 general manner. 



An imbibitionist might argue that since the normal direction 

 of the stream in wood is longitudinal, it naturally cannot be 

 transmitted by imbibition so easily in a transverse direction. We 

 have too little experimental knowledge of the nature of imbi- 

 bition to enable us to criticise this argument on general grounds. 

 But our experiments do yield an argument against the view in 

 question, though not one of great value. 



We have shown that on the average the depression in the rate 

 of absorption is considerably greater in Angiosperms than in 

 Gymnosperms. This at least seems to be true for Taxus as com- 

 pared with Angiosperms : the results obtained with Pinus being 

 possibly too discordant to be trusted. If on general principles it 

 is to be assumed that water of imbibition cannot travel easily in 

 the transverse direction we must assume it to be true for all kinds 

 of wood. There is no reason why it should be applicable to 

 Angiosperms and not to Gymnosperms. Therefore "double saw- 

 ing" ought according to the imbibitionists to produce the same 

 effect on Gymnosperms as on other trees — and this is not the 

 case. But it must be added that we are not able quite clearly 

 to explain the difference in question by any theory of water- 

 transport. 



It is clear, since double sawing destroys the vessels as 

 carriers of water, that therefore the operation in vascular plants 

 is equivalent to destruction of a large part of the water-channels. 

 And these channels are of large bore and without transverse 

 divisions, in other words the destroyed channels are those best 

 fitted for longitudinal transmission. But in wood made up of 

 tracheids as is that of Gymnosperms, this destruction of part of 

 the water-course does not take place. 



But it may be said that our experiments in which a single 

 deep cut was made, show that the destruction of a very great 

 part of the water-channel does not seriously depress the rate. 

 Hence it may be argued that the chief cause of the depression 

 in the double sawing experiments, is rather to be sought in the 

 fact that the current has to travel transversely than in the fact 

 that it has a diminished amount of tissue to travel in. 



But the above-mentioned remarkable fact (Exp. 23) that a 

 deep single cut does not seriously depress the rate shows that 

 transverse transmission is easily effected if there is given a suf- 



