8 E antes , — The Morphology of A gat his australis . 
occasionally in the mature egg, but their occurrence or position is not con- 
stant, as is the case in some Abietineae. The ripe egg nucleus may be 
found in any part of the cell, but its position is normally upper central. 
The germination of the pollen was not observed, but the behaviour of 
the tubes is even more unusual than reported by Thomson (15) in this 
genus and in Araucaria. The latter has described them passing along the 
cone scale, penetrating its tissue and that of the integuments. In Agathis 
australis the pollen-grain from its position in or near the axil of the 
scale sends tube-branches in various directions — into the nucellus and 
also away from the ovule into the cortex of the cone axis and into the 
basal portions of adjacent scales. The courses of those leading away from 
the ovule are very peculiar. Their direction is often apparently purposeless : 
some wander just beneath the epidermis and may pass to the micropyle of 
adjacent ovules ; others push directly inward, reaching the central cylinder 
of the cone. These in their course often follow the scale traces, running 
along or into the phloem, and even entering the xylem ! All the tissues of 
the cone may be penetrated ; only large resin canals prove obstacles. 
When a pollen-tube enters one of the latter it promptly retreats and takes 
a new direction. Penetration of the xylem of the scale trace by the pollen- 
tube is not constant in Agathis australis ; some cones show many instances, 
others none at all. The tube branches which behave in this strange way 
are usually small, and follow an irregular course — in and out from cortex to 
fibro-vascular bundle, dying out as tapering tips in either tissue. 
Among spiral and scalariform tracheides fragmentation occurs during 
the passage of the tube. Pitted xylem is probably only entered before 
lignification is complete. PI. IV, Fig. 41 shows a cross-section of a scale 
trace (within the cortex of the cone), the xylem of which has been entered 
by a pollen-tube. The latter seems to have penetrated while the xylem 
was in a plastic state, since the tracheides have arranged themselves smoothly 
about it. The shrunken pollen-tube wall is visible. The tube entered the 
trace from the side near the base of the cone scale, passed some distance 
through the phloem, entered the xylem, which it traversed for a few milli- 
metres, and returned to the cortex. Three other tubes are seen in the 
phloem on the lower side of the figure. 
This odd feature was observed in other species of Agathis during the 
study of the anatomy of the cone scale. It was not found in A. vitiensis , 
(Seem.) Benth. and Hook, f., perhaps because the cones examined were 
almost sterile. But the scale traces and even the axial cone bundles 
of A. alba , Rumph., and of A. Bidwillii (specific name as placed on material 
sent from Botanic Gardens, Buitenzorg, Java) show pollen-tubes traversing 
them, even more frequently than does A. australis , and those of A. borneen- 
sis , Warb., are very much eroded by them. In the latter species so 
serious is the attack in some cases that wonder is caused as to how the 
