April \, 1878] 



NATURE 



447 



produce fruit." ^ Perhaps therefore Voandzeia may have 

 to be expunged from the list of cleistogamic plants, while 

 on the other hand Krascheninikowia, according to a 

 thoughtful criticism of Mr. Darwin's book in the Journal 

 of Botany, must be restored to it. 



It may also be noted that according to Bentham Mar- 

 tinsia was a genus founded on a cleistogamic state of 

 Clitoria glycinoidis j Cologanta also should possibly be 

 added to the list since Zuccarini's Martia mexicana 

 appears to be an apetalous condition of some species of 

 the genus. 



Although the habit of producing cleistogamic flowers is 

 pretty widely diffused amongst flowering plants it is 

 locally concentrated in particular groups. This is par- 

 ticularly true in the case, as Mr. Darwin has pointed out, 

 of Malpighiacece and Acanthacece, and amongst Legu- 

 minoscB in the Glycinece. The genus Viola is remarkable 

 in this respect ; it is rich in cleistogamic species except 

 in the section Melanium, to which V. tricolor belongs. 

 In this species, besides conspicuous flowers adapted for 

 self-fertilisation, smaller and less conspicuous flowers 

 adapted for self-fertilisation are produced. These are not 

 closed, but, as Mr. Darwin points out, " they approach in 

 nature cleistogamic flowers," and though they diff'er in 

 being produced on distinct plants they are perhaps des- 

 tined to be as completely modified as the self-fertilising 

 flowers of other sections of the genus. 



The question as to the causes predisposing to the pro- 

 duction of cleistogamic flowers is one of very great inte- 

 rest. In the first place Mr. Darwin points out that the 

 larger proportion of known cases belong to plants with 

 irregular flowers, that is, to plants whose flowers have 

 been adapted for insect cross-fertilisation. Cleistogamy 

 in this light is a resource to fall back upon when the 

 elaborate adaptations for making insects do their work 

 fail, as they seem to do more or less in Viola. It is a 

 remarkable contrast that in heterostyled flowers, which 

 are absolutely dependent upon insects for their legitimate 

 fertilisation, irregular flowers are extremely exceptional, 

 the adaptation, as far as it goes, being so complete that 

 anything further in that direction is superfluous. 



Four cleistogamic genera are normally wind-fertilised, 

 and this shows that the cause alluded to above must be 

 a subordinate one. Mr. Darwin urges with much force as 

 the most potent agency, the unfavourable influence of 

 climatic changes. From the time of Linnaeus, it has been 

 observed that exotic plants may be fertile, though their 

 flowers have never attained proper expansion, that is to 

 say, for the nonce they have become cleistogamic and 

 self-fertile. The same thing occurs on a large scale with 

 Jtincus bufoniiis, in Russia, which in some districts never 

 bears perfect flowers, while in Liguria, Viola odorata never 

 bears cleistogamic ones. It is perhaps, however, doubtful 

 whether winter-flowering plants are absolutely sterile, 

 since the well-known Chimonanthus, whose name re- 

 cords its habit, is known to fruit, though sparingly, in 

 this country. The evidence is, however, strong enough 

 to render it highly probable that plants which are normally 

 cross-fertilised, are driven into the abasement of cleisto- 

 gamy when their geographical limits are extended beyond 

 the limits not favourable to their receiving visits from ap- 

 propriate insects, or to their properly expanding their flowers. 



* yoimu Litu Soi,, Eot xi. p. aSfl, 



Here our comments must cease, content for our part if 

 they attract a few more readers to a most fascinating 

 research. 



LETTERS TO THE EDITOR 

 \The Editor does not hold himself responsible for opinions expressed 

 by his correspondents. Neither can he undertake to return, 

 or to corresp07td with the writers of, rejected manuscripts. 

 No notice is taken of anonymous communications. 

 [ The Editor urgently requests correspondents to keep their letters a f 

 short as possible. The pressure on his space is so great that it 

 is impossible otherwise to ensure the appearance even of com- 

 munications containing interesting and novel facts.] 



Elements of Articulate Speech 



As a corollary to the interesting observation with the phono- 

 graph recorded by Prof. Fleeming Jenkia and Mr. Ewing in the 

 last number of Nature, will you allow me to point out that 

 every capital letter of the Greek alphabet except r and P is 

 actually (either as written or when turned through an angle of 

 90°) a reversible or a reduplicate symbol. 



With regard to gamma, although the capital is not, the soaall 

 letter (7) is reversible ; and as to P (or R, which is another 

 ancient Greek form of it), many facts seem to show that by 

 itself it does not as a rule represent a complete part of articu- 

 late speech ; witness its frequent reduplication in Greek, the 

 aspirate so often employed with it both in Greek and Latin, and 

 the way in which it is frequently omitted, as if of no importance, 

 from Latin words ordinarily spelt with it. The French or Italian 

 pronunciation of this letter amounts to a reduplication in the 

 English ear, while the English pronunciation of it amounts to 

 its omission altogether in the ear of a Frenchman, an Italian, or 

 a Scotchman. 



In the Roman alphabet F, G, L, P, and R, are the excep- 

 tions ; much might be said about each of these, but I will con- 

 tent myself by saying that L is obviously only an apparent 

 exception, as it is easily derived from A. W. H. Corfield 



10, Bolton Row, Mayfair, March 30 



Phoneidoscopic Representation of Vowels and 

 Diphthongs 



I HAVE just obtained the two following results with the 

 phoneidoscope ^ : — 



1. If a vowel be steadily sung on a single note, a constant 

 colour- figure is produced ; but if the vowel be spoken in the 

 ordinary conversational tone, a change of figure occurs before 

 the sound ceases. The slurring alteration of pitch which takes 

 place in pronouncing a single vowel is thus rendered perceptible 

 loy the eye. 



2. When a diphthong is slowly intoned, two distinct figures 

 successively present themselves, which are found on trial to be 

 those corresponding to its constituent vowel-sounds. The two- 

 fold nature asserted in the word "diphthong" receives by this 

 experiment d. vidble illustration. Sedley Taylor 



Trinity College, Cambridge, April I 



The Southern Drought 



You ask in last week's Nature (p. 436) for information 

 respecting the drought in the southern hemisphere. A few days 

 ago I received letters from Samoa and the Gilbert Islands telling 

 me of its severity there. Droughts are of frequent occurrence in 

 the Gilbert Islands, but my correspondent (a native of Samoa) 

 tells me they have had an extraordinary one there, which com- 

 menced in 1876, and which continued up to the date of his 

 letter — December 4, 1877. lie says many of the peopls have 

 died from starvation in consequence. 



A letter from a missionary who has been forty years in Samoa 

 contains the following : — " We have had the greatest drought I 

 have ever known." The Samoan Islands are wonderfully fertile, 

 and even during what is called the dry season it is rarely that 

 more than a fortnight passes without rain. The atmosphere is 

 always full of moisture, and there are very heavy dews at night, 

 so that the vegetation never gets burnt up, except the drought 

 be very extraordinary. Now, however, my correspondents speak 

 of scarcity of food in those most fertile islands. 



Blackheath, March 29 S. J. "Whitmee 



[Can our correspondent favour us with the date of the last 

 drought or series of droughts? — Ed.] 



' See Nature, vo' .wii. p. 476, note 2. 



