SCIENCE-GOSSIP. 



53 



The degree of coincidence is so generally ap- 

 proximate all along the line that it seems probable 

 the minor deviations will disappear when the values 

 of the divisions have been more accurately deter- 

 mined ; and it seems to be quite within the limits 

 of the possible that the one or two cases of entire 

 lack of coincidence may yet be reconcilable. The 

 case of non-coincidence here alluded to is the 

 occurrence of mammals in many parts of the 

 secondary strata from those of the Triassic period 



upwards. This point need not be discussed here, 

 but some reference to it is necessary because an 

 unquestionable lack of coincidence of these scales 

 during any period prohibits its application to the 

 Cambrian and all other periods. 



Such is a rough outline of the idea of time- 

 valuing from a general point of view. I propose 

 to apply it in two or three ways to the Lower 

 Cambrian period, in the subsequent part of this 

 paper. (To be concluded next month.) 



POLLINIA OF BRITISH ORCHIDACE^. 

 By H. E. Griset. 



■HpHE androecium of the Orchidaceae is un- 

 doubtedly a very peculiar and anomalous 

 structure, something analogous occurring in the 

 Asclepiadaceae, a dicotyledonous order of which 

 there are no British representatives. In the 

 Orchidaceas the flowers are trimerous, hence the 

 androecium previously consisted of three parts, one 

 stamen superior, that is opposite the labellum, and 

 two lateral, their three filaments being confluent or 

 adherent with the style, forming collectively the 

 column or gynastemium, the androecium being said 



by a glandular pulvinate body called the re- 

 tinaculum ; the retinacula are free or connate, or 

 may be contained in the same or in distinct pouches ; 

 the pollen is commonly collected into roundish 

 yellow corpuscules, which are held together by 

 minute, viscid, elastic threads ; these converge 

 below into, and form the caudicle. This short 

 description of the relative position and general 

 structure of the androecium is given for the con- 

 venience of the general or inexperienced botanical 

 reader. I will now proceed to describe in detail 



to be gynandrous or united to the style. One — the 

 superior one — is usually perfect, and the two lateral 

 ones abortive. In Cypripcdium the reverse of this 

 occurs, that is the superior one is abortive while 

 the two lateral ones are perfect. The anther or 

 anthers thus appear sessile as it were, and inserted 

 upon the apex of the column, or rostellum of the 

 stigma ; the anthers are two-celled, and their apices 

 are sometimes elongated into a 'beak.' Each cell 

 contains a pollinium or pollen-mass, which may be 

 either sessile or surmounted upon a pedicel or 

 caudicle ; this latter is generally terminated below 



the minute and characteristic peculiarities of 



several of the common and more typical species. 



These pollinia may be divided into two types, 



those pedicellate, or having a caudicle, and those 



that are sessile, or have no caudicle ; these 



peculiarities together with those of the anther, etc., 



form valuable materials for the classification of the 



species, those above being in strict conformity with 



the sub-divisions of the order of Hooker, the 



former being that found in the tribe Ophrydae, and 



the latter that of the Neottideae of his 



classification.* 



Student's Flora. 



