an archaic type of Seed from the Palaeozoic Rocks. 103 
which collectively form the integument is susceptible of the most diverse 
interpretation. Two views have already been put forward, and as these are 
fundamentally opposed, and illustrate quite distinct tendencies in morpho- 
logical interpretations, the matter is of some little interest . 1 Many 
morphologists hold that all new structures are really fashioned out of old 
ones that have undergone a functional change. The petal is a sterilized 
sporophyll ; the paraphyses of mosses are sterilized antheridia ; the cortex 
of the aecidial fruit in the Uredineae is the reduced product of fertile 
hyphae. Of this order is Miss Benson’s suggestion that the integument 
of Physostoma has been elaborated by the sterilization of the peripheral 
sporangia of an ancestral synangium, the central member of which is repre- 
sented by the nucellus of the seed — a suggestion which is embodied in her 
synangial theory of the seed. That sporangia may undergo sterilization 
and persist as paraphyses scattered through the sori is illustrated by many 
present-day ferns, more particularly by representatives of the Polypodiaceae, 
as in Polypodium verrucosum , Vittaria rigida , V. Forbesii , and V. angusti - 
folia , Acrostichum aureum , &c . 2 3 Among palaeozoic Protofilices somewhat 
similar structures have been attributed by Renault to Boiryopteris forensisP 
In the absence of special investigations elucidating the nature of these 
curious structures, it may be admitted provisionally that their sporangial 
derivation seems probable, and that their existence gives plausibility to the 
synangial theory. 
When we turn to cases offering a closer parallel to the seed condition, 
such as Lepidocarpon and the megasporocarp of Azolla , the facts that have 
been ascertained are less favourable to the theory. In Lepidocarpon it is 
difficult to interpret the integument as other than a special production or 
enation of the supporting bract or sporophyll, whilst in Azolla the non- 
functional sporangia, which are present immediately below the mega- 
sporangium, are found to abort during development, whilst the indusium 
or sporocarp-wall arises as an independent annular upgrowth below the 
point of insertion . 4 
Cases such as these show clearly enough that the possibility of 
integuments arising as new or special formations (where seeds and similar 
structures are involved) cannot be lightly dismissed. The synangial theory 
though no doubt tenable, is, after all, no more than a hypothesis, which pre- 
supposes in the ancestor the existence of the exceptional condition of 
a synangium in which the peripheral members were ranged symmetrically 
around a central sporangium, and in which — unlike Azolla — they persist as 
a sterilized envelope to form the seed-coat. 
1 Miss M. Benson, loc. cit., p. 161 ; Oliver and Scott, loc. cit., p. 232. 
2 See Hooker’s Genera Filicum, Tab. 14, 68 B, 76 B, 77 A, 81 A, &c. 
3 See Renault, Bassin houiller et permien d’Autun et d’Epinac. Flore fossile, pt. 2, p. 54. 
4 See Goebel’s Organography of Plants, Pt. 2, p. 488, Fig. 325. 
