78 Mr. J. Miers on the Tribe CoUetiese, 



connected with their structure^ those of Colletia dumosa, collected 

 by me in Chile thirty-five years ago, are selected as an exam- 

 ple. They are of an oval form, about 1| line long, 1 line in 

 diameter, sometimes slightly compressed, often with an obsolete 

 ridge on the ventral face, which has been mistaken for a raphe ; 

 erect ; with a very hard polished outer surface, appearing finely 

 granulated w^hen magnified : this outer coating is corneous in 

 texture, becoming softer by long maceration, when it may be 

 cut through with a sharp knife ; it is somewhat translucent, but 

 rendered densely black by the lining that adheres to it ; at its 

 base, in a transverse direction in regard to the fruit, is seen a 

 long furrow, or compressed bilabiate slit, which is always per- 

 vious. This outer shell is uniform in thickness, homogeneous in 

 texture throughout, and under the microscope is seen to be com- 

 posed of numerous narrow, transverse, hyaline cylinders, appa- 

 rently solid, of equal size, all placed in a direction radiating from 

 the centre of the seed ; so that when the shell is viewed through 

 either of its faces, it appears like a reticulation of minute hexa- 

 gonal cells, each cell corresponding to the end of a long cylinder : 

 these cylinders, by long maceration, may be separated from each 

 other ; it is clear, therefore, that in this organization thei'e is 

 not the slightest approach to any longitudinal vascular conti- 

 nuity. For reasons to be shown presently, we may infer that 

 this shell is not one of the ordinary integuments developed from 

 any of the original tunics of the ovule, but of posterior growth, 

 and therefore not a testa, as hitherto supposed : it is a produc- 

 tion apparently analogous in its structure to the outer tunic 

 formerly described in the seeds of Cucurbitacem^. 



* Trans. Linn. Soc. .xxii. .92 & 108. More extended observations upon 

 the seeds of different Cucurbitacece have led me to modify the inference 

 made in that stage of the inquiry {loc. cit.), that the more external covering 

 (there described as formed of three several portions) consists of epiderm, 

 mesoderm, and endoderm — terms incorrectly applied to parts evidently 

 distinct in their origin. The outer very thin stratum of that fleshy cover- 

 ing is reticulated with long hexagonoid cells, and its origin may be traced 

 to an extension of the outer membrane of the funicle ; the intermediate 

 soft sid)stance is a collection of lax fleshy cells intermixed with numerous 

 araneiform filaments, all probably emanating from the fleshy substance of 

 the funicle. In many geiiera this fleshy ])arenchymatous substance is en- 

 tirely wanting, so that the shell of the seed is covered simply by the above- 

 mentioned thin pellicular membrane. In the centre of the funicle a thick 

 cord of vessels is seen, which, as it ajjproaches the seed, is observed to 

 bend towards one side, and to enter one of the ends of the linear basal slit 

 of the hard tunic ; there is no appearance of its bifurcation at that ])oint, as 

 the vessels of the funicle appear to be iumiediately connected with only 

 one of the two extremities of the peripherical raphe, which meet together 

 in the base of that tunic. This hard crustaceous shell, usually called the 

 testa, presents no sign of any organic connexion with the two layers of 

 structure before mentioned, nor with the more internal tunic. The exist- 



