2 54 BOTANICAL GAZETTE [octorer 



J. 



gonium complex. It is not separated from the egg by a wall, 

 but lies free in the cytoplasm of the egg. The figures of Stras- 

 burger (ii) and Belajeff (3) show that it takes a similar position 

 in Juniperus, 



Arnoldi (2) claims that there is no ventral canal cell in the 

 Cupresseae, and among the Taxodieae he failed to find it in 

 Cunninghamia, Sequoia, Cryptomeria, and Taxodium. Coker 

 (5), however, finds a ventral nucleus in Taxodium, although it is 



not separated from the ^gg by a wall, lying free in the cytoplasm 



of the ^%%* He also finds the ventral nucleus very prominent in 

 Podocarpus, but here also no wall is formed. Strasburger (iij 

 as long ago as 1879 figured the spindle in Juniperus. There 

 can be no doubt that the free ventral nucleus without a separat- 

 ing wall is a specialized condition. A similar reduction in the 

 neck canal cells can be seen as one traces a series through the 

 bryophytes and pteridophytes. In most of bryophytes the 

 number of neck canal cells is rather large; in the pteridophytes 

 the number has become small, often only two, and even here a 

 wall seldom forms between the two nuclei. In most gymnosperms 

 there is a definite ventral canal cell, cut off by a conspicuous 

 wall ; in Taxodium, Podocarpus, and Thuja the nucleus is formed, 

 but the wall is suppressed. This may be the case in the forms 

 described by Arnoldi (i, 2), and may have led him to think that 

 the ventral canal cell was not represented. We doubt whether 

 a ventral canal cell is lacking in any gymnosperm, and believe 

 that the absence of such a cell, or at least a nucleus representing 

 it, could be established only by the most indubitable evidence. 



When an archegonium has not been fertilized, there are indi- 

 cations that the ventral nucleus divides {Jig. ij). Coker (5) 

 holds that in Podocarpus these divisions are amitotic. In Thuja 

 after fertilization there are frequent instances of a further 

 division of the ventral nucleus, and these divisions are mitotic. 

 Coulter and Chamberlain (7) suggest that in some cases the 



ventral nucleus in Pinus may be fertilized instead of the cgg^ A 



number of the writer's preparations of Thuja lead him to believe 

 that both the ventral nucleus and the eggr in the same archego- 



