APPENDIX 575 



occurred to me that this seed-structure might be the result ol a lost vivi 

 parous habit. One apparently had to deal here not with an ordinary si 

 containing an embryo in the midst of albumen, but with a seed in an 

 arrested stage of germination surrounded by a body that might peril 

 prove homologous with the " cotyledonary body " of Rhizophora. Th< 

 process of development that goes on without a break in Rhizophora, from 

 the fertilisation of the ovule to the detachment of the seedling from th< 

 branch, was here, as I considered, arrested after germination had begun, 

 but before the protrusion of the seedling from the fruit. With nearly all 

 plants, as I reflected, there is a rest-stage of varying length, which might b< 

 called the seed-stage. With the mangrove-genera, Rhizophora and Brugui* 

 I had convinced myself by a long series of observations, the results of which 

 are given in Chapter XXX., that this rest-stage does not exist. It occurs 

 I argued, in Barringtonia, but only after germination has begun, and, there- 

 fore, displaced when compared with the typical seed-stage of most plan; 



In this connection it may be noted that a difference in germinal 

 behaviour might be expected between the two shore species on account 

 their difference in stations, Barringtonia speciosa growing on the sands 

 beach, and B. racemosa in the wet ground around a mangrove-swamp. 

 There is a strong suspicion that the rest-stage in B. racemosa is very short, 

 though I never found germination in progress on a tree (see Note 3; 

 There is no doubt, on the other hand, that the rest-stage of B. speciosa is 

 often, as with most other plants, very long. This, then, was my lesson 

 from the Barringtonia fruits on the banks of the Rewa, and the question 

 arose whether this interpretation of these curious seed-structures accord 

 with the opinion formed of their nature by botanists. 



Curious seed-structures of this kind must have their significance in the 

 history of the plant ; and on returning to England I looked a little further 

 into the matter. To follow up this kind of inquiry, however, would carry 

 me far beyond the limits prescribed for this note, and I have only b 

 it here in a tentative fashion. Different botanists of eminence have paid 

 attention to this subject, amongst them Roxburgh, Thomson, and M 

 (see Dr. T. Thomson in Journ. Linn. Soc. Bot., vol. ii., p. 47, 1858, and 

 Mr. J. Miers in Trans. Linn. Soc. Bot., vol. i., 1SS0). It would appi 

 that the seed-structure of Barringtonia is also found in ( 'areva, a genus 

 the same Myrtaceous tribe, and in Garcinia and other genera of the 

 Guttiferas, as well as in other inland plants. 



Mr. Miers, after reviewing the opinions of his pivd 

 results of his own investigations. The solid embryo found in Barringtonia 

 and many other genera consists, he observes, (a) of an external p< 

 the "exorhiza," which nourishes the germinating seed and then dii - awa 

 (b) of an internal portion, the "neorhi/a," which, growing a! each end, 

 forms the central portion of the stem and root ; and (c) the "medull 

 sheath " of Mirbel, that lies between the two, and is composed of el< 

 tary vascular tissue, which ultimately gives origin to the wood, hark, and 



