308 BOTANICAL GAZETTE [October 



young strobilus is one with the bract and becomes evident in the 

 later stages by the comparatively rapid growth of the tissues on 

 the dorsal side of the composite organ. 



In the fourth group there is no external evidence of more than 

 one organ. Into this group fall Arthrotaxis sclaginoides (i), 

 Agathis (fig. 169), Saxegothaea (5, 8), Pliyllocladus (fig. 183), 

 Taxus (fig. 191), Torreya (6, 7), and Ccphalotaxus (6, 7). It should 

 not be surprising to find forms in which the welding has taken 

 place beyond the recognition of more than a single structure when 

 one considers to what extent this process has taken place in Chamae- 

 cyparis, Junipcrus, Thuja, Cunninghamia, and Podocarpus dacry- 

 dioidcs (4). The low cushion behind the ovule in Agathis 

 australis suggests the complete fusion of a scale to a large bract; 

 a similar fusion is nearing its completion in Cunninghamia 

 Davidiana. 



On the basis of vascular anatomy the investigated sporophylls 

 fall into two general groups. 



In the first group the bract and scale supply arises as separate 

 bundles in the cylinder gap. In this group belong in general those 

 forms in which the two sporophyll parts are separate and fairly 

 well developed, as the seed-producing sporophylls of Finns Kete- 

 Iceria, Picea, Larix, Tsuga, Pseudotsuga, and Abies (2). To this 

 group belong also many in which the two sporophyll members 

 present considerable to complete fusion, as Araucaria Bidwilli 

 (1, 3, 10), Chamaecyparis Lawsoniana, Junipcrus communis, and 

 the upper sporophylls of Thuja occidentalis, Cupressus Bcnthamii, 

 and Cryptomeria japonica. This group includes also some in which 

 the sporophyll is evident only as a single organ, namely Pliyllocladus 

 and Ccphalotaxus (6, 7). 



In Podocarpus and Dacrydium, where the strobilus consists of 

 one or two sporophylls, and in Junipcrus communis, the sporophylls 

 receive the final bundles of the axis. There is in these instances no 

 cylinder gap, and the bract and scale supplies, at least in the forms 

 investigated by the writer, result from the division of one of the 

 final bundles in the axis. The early division of the bundle in the 

 tip of the axis perhaps justifies the placing of these forms in this 

 group. 



