October 1, 1898.] 



KNOWLEDGE. 



223 



opening by two valves ; and the seeds contain no reserve 

 material or endosperm. 



The flower of the bladderwort is particularly remarkable 

 for the extreme irritability and sensitiveness of its stigma ; 

 the two lobes close together immediately on being touched, 

 but open again after two or three minutes if no pollen 

 grains happen to be enclosed in their embrace. A precisely 

 similar device is to be found in the yellow monkey-tlower 

 (Miuiulus lutius) which is sometimes found floating in 

 golden masses on still and silent pools. A flower-haunting 

 fly, such as one of the hovering Syrjiltida, will alight on 

 the lower lip of the corolla, and in thrusting his proboscis 

 down the tube in order to reach the honey in the spur, 

 will first of all rub his back against the stigmatic lobes 

 which project beyond the anthers. Directly afterwards he 

 wiU be dusted with fresh pollen and will be ready to carry 

 it to the nest flower he visits. The sensitive folding 

 together of the stigmatic lobes is thus a safeguard against 

 self-fertilization, for when the insect, laden with pollen, 

 withdraws from the flower, the lobes will have their receptive 

 surfaces in close contact with each other. Yet, if the 

 flower is not fortunate enough to secure the advantages of 

 cross-fertilization by insect agency, it will take to self- 

 fertilization as a last resource, the stigma curling round 

 backwards so as to receive the pollen which at first it was 

 so careful to avoid. 



The butterwort (Pinfjuicula) is in some respects more 

 highly specialized than its cousin the bladderwort — at least 

 from a physiological point of view — for its leaves can not 

 only catch insects with a greasy sticky secretion, but can 

 also digest them (just as in the sundew) by means of the 

 ferment pepsin. Moreover, the flowers of the butterwort 

 reach a higher note in the scale of colour ; Pinguicula alpina, 

 indeed, is yellow, and is fertilized by flies (Syrp/iidce), but 

 P. fuhjaris and P. ijraitdijlora are deep blue and adapted 

 for bees. '' 



The butterwort on the other hand is provided with roots, 

 and still shows the primitive characteristic of a rosette of 

 simple undivided leaves, of which only a trace exists in the 

 youngest stage of the bladderwort, although this rosette is 

 more noticeable in the terrestrial species. The stigma of the 

 butterwort does not show auy sensitiveness to the touch ; 

 its lower lobe merely hangs down like a curtain in front 

 of the anthers so as to intercept any pollen which may 

 be brought by a winged visitor. Self-fertilization, how- 

 ever, may likewise occur if no pollen has been transferred 

 by insects from other flowers. Finally, the embryo of 

 the butterwort has not reached quite so low a state of 

 degeneracy as in the bladderwort, because it is still pro- 

 vided with a seed leaf. 



We have to turn to the tropics in order to find 

 another member of the order, which will show inter- 

 mediate characteristics between our bladderwort and butter- 

 wort. This is the genus irenlisea of Brazil, which retains 

 the primary rosette of leaves : the stem is, however, 

 thickly covered not only with unmodified spatulate leaves, 

 but with others metamorphosed into curious insect traps, 

 long-necked bladders with a kind of spiral entrance, thickly 

 beset with hairs, pointing backwards and preventing any 

 escape. Genlisea is a land plant, but agrees with the 

 bladderwort in being destitute of roots. 



The tropics, again, are the home of the Gesneriacefe, the 

 order with which the Lentibulariacere show the closest 

 genetic relationship, for although the flowers possess 

 external resemblances to some of the more distant 



Scrophulariacece, such as snapdragon and calceolaria, yet 

 these similarities are no more than what all three orders 

 possess in common. 



The Gesneriaceir form a family well known to gardeners 

 for the handsome and showy flowers comprised within its 



2. Internal ab- 

 sorptive hairs, 

 still further 

 magnified. 



1. Bladder of Bladderwort, in dia- 

 grammatic section, magnified. 



limits, such as tlesnera. Gloxinia, Achimenes, .Eschynan- 

 thus, etc. 



The Lentibulariacefe possess so many points in common 

 with the Gesneriaoese, especially with the subdivision 

 Cyrtandreiie that they might almost be classed with the 

 latter ; thus, in both oases the ovary is unilocular and the 

 seeds are without endosperm. Moreover, many of the 

 Cyrtandrese have only two stamens, and the genus Strepto- 

 riirpns agrees with Utricuhiria in the absence of even a 

 primary root. 



It may be added that the parasitic family of broom- 

 rapes (Orobanchaceie) is also closely allied to the 

 Gesneriacene, and agree, too, with the bladderwort in the 

 embryo being totally undifferentiated, consisting merely of 

 an oval cellular mass. 



In conclusion, the different relationships maybe graphi- 

 oally represented thus : — 



Personatie. 



I 



^1. I . I 



Cresneriaceae. Scrophulariacese. Solanacese. 



Cvrtandreae. Gesnerieee. 



Orobanchacete 

 (parasitic). 



Lentibulariacese 

 (insectivorous). 



• The small Lusitanian butterwort is, however, pale Ulac in colour, 

 and depends only on self-fertilization. This is, perhaps, a case of 

 reversion from the blue flower fertilized by bees. 



ETHNOLOGY AT THE BRITISH MUSEUM. 



By R. LVDEKKER. 



SINCE, so far at least as his bodily structure is con- 

 cerned, man evidently forms but the highest develop- 

 ment of the mammalian type, it is evident in every 

 well-arranged museum he should take his proper 

 position at the head of the series, adjacent to the 

 man-like apes. And it is therefore in the highest degree 

 satisfactory that this has at length been recognized by the 

 authorities of the natural history branch of the British 

 Museum, where an ethnological series is now in process 

 of formation and arrangement in the upper mammalian 

 gallery. It is not, indeed, that this is an entirely new 

 departure, for ever since the transference of the natural 

 history collections from Bloomsbury to South Kensington, 

 human skuUs and skeletons were arranged in serial 

 position in the gallery of osteology, which formerly occu- 

 pied the whole of what is now the upper mammalian 



