August 23, 1906] 



NATURE 



43' 



pyle closed in, the conservative nucellus would for a while 

 endeavour to maintain direct communication with the 

 exterior. The beak-like appendage on this view would be 

 a new formation evolved fari passu with the integument. 



A peculiar and distinctive, though negative, feature 

 common to the whole range of Pal;i;ozoic seeds that have 

 become known to us is the lack of an embryo. Occasion- 

 ally small-sized seeds are met with, as in Lagenostoma 

 Lomaxi, and now and then immature-looking stages, of 

 which the best example is Renault's Cordaitcan ovule, so 

 often figured in the books. But apart from such rarities 

 the petrifactions agree in being at a stage which, in the 

 light of recent Cycads, is to be interpreted as corresponding 

 lo the time of fertilisation. The pollen-chamber is charged 

 with pollen-grains, whilst in good examples the megaspore 

 is filled with a prothallus which frecjuently shows indica- 

 tions of archegonia at its upper extremity. All these 

 specimens will be dismissed by some as abortive, and any 

 conclusions drawn from the negative character as invalid. 

 Without ignoring this contingency another view is, of 

 course, possible. The normal fall of the seed may have 

 followed pollination at a short interval, much as is reported 

 for Cycas and Ginkgo to-day. The " resting period " in 

 these seeds would then perhaps coincide with the matura- 

 tion of the sperms, whilst the subsequent embryonic history 

 might have been carried through without a pause. This 

 view gains support from the filicinean relationship, for of 

 course the fertilised egg of a Fern continues its develop- 

 ment without interruption. If the modification of the 

 pteridophytic life-history that culminated in these early 

 seeds were directed, as seems probable, to ensuring a 

 greater certainty in bringing the gametes together under 

 conditions favourable to their union, it would follow that 

 the other great advantage arising from the seed-habit was 

 of later acquisition. In other words, the ordinary seed 

 with resting embryo was evolved by stages. There is a 

 great lacuna in our knowledge of the early adjustment of 

 the embryo to intraseminal existence. Whilst evidence of 

 Palceozoic seeds with resting embryos is altogether want- 

 ing, we are confronted in the Mesozoic rocks with the 

 Bennettitese, all of which possess a well-marked dicotyle- 

 donous embryo practically filling the seed-cavity. It is 

 mere conjecture to suggest that this change has been 

 wrought in response to some climatic stimulus, though the 

 marked xerophilous facies of many of the Mesozoic Cycado- 

 phyta seems quite consistent with such a view. Be that 

 as it may, one cannot fail to recognise that the resting seed 

 with an embryo marks a great advance on the Pterido- 

 sperm, an advance hardly less important to the welfare of 

 the plant than was the earlier type of seed on the extended 

 life-history of the filicinean prototype. 



This stage of the seed-history would be of exceptional 

 interest if we could hope to recover any morsels of direct 

 evidence. As yet we remain in the dark as to the morpho- 

 logical nature of the embryonic organs, how far we arc 

 dealing with new structures produced from a protocorm, 

 as Prof. Bayley Balfour has suggested ; ' how far they 

 represent the old filicinean organs adjusted to intraseminal 

 life. What chance there may be of the solution of this 

 difficult problem by the application of other methods may 

 emerge perhaps from the discussion on the phylogenetic 

 value of early seedling characters which is to be opened 

 next Tuesday morning by my colleagues Mr. Tansley and 

 Miss Thomas. 



Reference has already been made to the view that the 

 seed, as we find it in the majority of spermophytes with 

 its resting embryo, shows definite adaptation to seasonal 

 periodicity. It would be interesting to learn how far the 

 seeds of plants long accustomed to uniform conditions, 

 such as the rainy tropical forest, behave in this respect. 

 The point does not appear to have been verv fully investi- 

 gated. Indeed, there is a rich field for both observational 

 and experimental work upon obscure seed-problems await- 

 ing any one who can devote continuous attention to the 

 subject. Is there any solid foundation for the supposed 

 " physiological dimorphism " among seeds according to 

 which, as one reads in the older books, the earlier ripening 

 seeds are adapted to an immediate germination, whilst the 

 later ones are reserved for the follow-ing spring? It may 



1 Presidential Address, Section K, Glasgow, igoi, p. 9. 



NO. 192 1, VOL. 74] 



be that we have here but one more illustration of the 

 operation of temperature as the limiting factor, but in any 

 case the matter wants clearing up. An experimental in- 

 vestigation of the relations of " albuminous " and " e\- 

 albuminous " seeds would probably repay the trouble 

 involved. Does any condition or set of conditions under 

 the control of the operator exert an influence in this con- 

 nection? 



The mention of the early germination of seeds brings to 

 mind the most striking instance of all — that of the tropical 

 Mangrove, in which, as is so well known, the seed 

 germinates on the tree, so that the young plant is extruded, 

 and in some instances falls, from the parent free of its 

 envelopes. 



Our interest in this type of vegetation has been revived 

 through the researches of Mr. H. B. Guppy incorporated 

 in his recent contribution on " Plant-dispersal in the 

 Pacific." This volume, perhaps the most important con- 

 tribufion to the biology of tropical plants that has appeared 

 since the death of the lamented Schimper, is distinguished 

 alike for its wealth of new observations and its engaging 

 freshness of treatment. There is one suggestion of Mr. 

 Guppy's concerning the vivipary of Mangroves which may 

 occupy our attention for a few moments. 



As a result of his studies in the Pacific and elsewhere 

 Mr. Guppy has arrived at the conclusion that the Man- 

 grove type of vegetation is a very ancient one, dating bark 

 to the times when climate was more uniform and moist 

 than we know it to-day. The viviparous habit he cor- 

 jectures to have been once very general, whilst to-day this 

 primitive condition is making its last stand along thi' 

 tropical shores. Traces of vivipary still occur among 

 inland plants, such as Crinum, whilst in other cases it re- 

 appears intermittently under conditions not fully ascer- 

 tained. Mr. Guppy supposes the ordinary fruiting way of 

 plants with caducous fruits or seeds, that germinate after 

 an interval, to have arisen by a modification of the con- 

 tinuous viviparous method in the sense that the seed has 

 come to fall earlier and earlier until the stage now 

 characteristic of practically all Spermophytes has been 

 reached. 



Piecing the data together, this seems to be the position : 

 The earliest known seeds appear to have remained on ihf' 

 plant just long enough to receive their pollen ; but in time, 

 it is reasonable to suppose, the advantage of remaining 

 longer was realised, and the fall of the seed was postponed 

 until fertilisation was followed by the occupation of the 

 seed-cavity by an embryo. Here in seclusion the embryn 

 could remain until germination was convenient. Starting 

 at the other end. our modern seed, according to Mr. Guppy, 

 h.as been evolved by the gradual retention of the viviparou-; 

 embryo ; or, to put it in another way, the detachment of 

 the seed has been hastened so that it falls long before 

 germination is due. 



Well, these theories fail to meet in the middle, as they 

 should if they are to present us with an epitome of the 

 whole seed-history. Perhaps there were troublous times in 

 that middle epoch, so that the continuity has become 

 obscure ! Or possibly another view may be admissible of 

 the relation of vivipary to normal seed-production. Most 

 botanists, I take it, have been inclined to regard vivipary 

 as the dernier cri in seed-history, the ultimate stage in 

 the way of possible reproductive advance in seed-bearing 

 methods that the higher plants have yet attained. The 

 Mangrove process might even be conceived as the starting- 

 point, under certain contingencies, of a whole new race of 

 plants with life-histories complicated by fresh alternations 

 — homologous alternations — far beyond any of which we 

 have knowledge to-day ! 



Schimper and others who have given attention to the 

 subject found no reason for regarding vivipary as other 

 than an adaptation to special circumstances, an extreme 

 condition that had arisen independently in several cycles 

 of affinity. Before the contrary can be accepted a good 

 deal of positive evidence will be needed, drawn from the 

 non-Mangrove representatives of groups in which vivipary 

 occurs, to show that the relationship is other than has 

 been generally supposed. Moreover, if the viviparous habit 

 were formerly of wide occurrence some traces of it might 

 reasonably be expected in the fossil record. So far as can 



