0BSEBVATI0NS UPON THE POLLEN TUBE. 203 



seems always to be, or to become, more or less dicho- 

 tomous. It may be dichot odious from the commeDcemeDt 

 but in other cases the braDch commences as a swelling a 

 little below the apex, and is hence originally lateral, 

 afterwards lengthening so rapidly as to rival the length of 

 the original apical portion and hence becoming dichoto- 

 mous. It is rarely that a tube branches more than once 

 and I have never seen more than two branchings occur 

 in the same tube. 



Pollen graius of Mercurialis may emit two or eveD three 

 separate tubes all arising close together but distinct to 

 their bases. In other cases three tubes are formed by the 

 pollen grain forming the primary tubes and one of these 

 branching dichotomously. The formation of two tubes 

 from one grain is of fairly frequent occurrence but these 

 tubes are only very rarely seen. The formation of cellu- 

 lose ingrowths is in the case of Narcissus most frequent 

 is sugary solutions of some strength whereas since the 

 ingrowths are really irregular internal swellings produced 

 by the excessive localized imbibition of water, we should 

 expect to find them more abundantly formed in weaker 

 than in strong sugar solutions. On the other hand the 

 percentage of solids in the fluid, coutained in the pollen 

 tube and touching the inner side of the wall, will be less 

 as compared with that of the external medium when in 

 a strong, than when in a weak, sugary, solution. Hence 

 at any spot where the sugar penetrates in less relative 

 proportion to the water that its does elsewhere and where 

 the internal layers of the wall of pollen tube are soft and 

 readily capable of imbibition an internal swelliug in the 

 form of an irregular cellulose ingrowth is produced. 



It sometimes happens that wrinklings of the wall of the 

 tube simulate an appearance of septa. These form thin 

 lines running generally diagonally across the tubes and 



