LIFE HISTORY OF PINUS 1 39 



until it came into contact with the agg, when an entrance was 

 effected through the sheath-cells. That this has cost the pollen- 

 tube an unusual effort would seem to be evidenced by the fact 

 that it has become filled with a cytoplasm as dense as that of 

 the egg, whereas, normally, its cytoplasm is very scanty. If 

 it be true, as Lidforss ('99) claims, that the penetration of the 

 pollen-tube is simply due to a search for food, it would appear, 

 in such a case as this, that the pollen-tube is capable of very 

 intelligent searching. In this instance the relation of the pro- 

 thallium to the nucellar cap is very like that found in the more 

 primitive Gymnosperms such as Zamt'a, Cycas and Ginkgo. 

 Here, however, the sperm-cells being non-motile it was neces- 

 sary, if fertilization take place at all, that the pollen-tube should 

 reach the egg. 



The variations recorded here have a certain interest both 

 phylogenetically and ontogenetically ; but the most significant 

 lesson to be derived from them is the warning that they sound 

 against basing conclusions on meager observations. When 

 this is done, misconceptions and actual errors are bound to be 

 promulgated for truth. 



Botanical Department, Wellesley College, Dec. 28, 1902. 



NOTE. 



This paper was completed on December 28, 1902. During 

 the time that has since elapsed much valuable literature dealing 

 with subjects more or less intimately connected with questions 

 herein discussed has appeared. 



It is not feasible to make adequate references at this time to 

 all these papers, but the more important ones are mentioned be- 

 low, and the page in this paper, where mention of the views 

 recently expressed by other writers should have appeared, is 

 indicated. 



The first 88 pages of this paper were printed before the writer 

 was able to obtain copies of some of the articles mentioned be- 

 low. I regret, therefore, that it was not possible in many in- 

 stances to refer by means of a foot-note to the references made 

 in this addendum. 



Page 22. Strasburger ("04) now believes that synapsis is 



