August, 1911. 



KNOWLEDGE. 



317 



leaves I collars I at its apex, and when there are more than two 

 ovules on a stalk we find twice as many bundles in the stalU 

 as there are ovules at its apex — just as there should be if each 

 collar were a carpel. 



The development of the microsporangium (pollen-sac) has 

 recently been described by Miss Starr, and it mature structure 

 and mode of dehiscence by Goebel. In development, 

 apparently a single hypodermal cell divides to form the tapetal 

 and sporogenous tissue, and. outside of the tapetum, the wall 

 of the sac (except the outermost or epidermis layer). The 

 wall consists of from four to seven layers of cells, the hypo- 

 dermal layer and the layer below it being thickened by fibres 

 serving for dehiscence. Goebel i Flora, 1902) has studied the 

 structure of the wall of the ripe pollen-sac ; apparently Ginkgo 

 is the only gymnosperm that has an endotheciuni, the 

 dehiscence layer or layers being of hypodermal origin ; the 

 longitudinal slits of dehiscence of the two pollen-sacs face 

 each other, and lie at such an angle that the pollen is readily 

 shed for wind dispersal. 



The development of the ovule has recently been worked out 

 by Miss Carothers {Bot. Oaz.. 1907), who extends the earlier 

 accounts. The ovule resembles in general structure those of 

 cycads and cordaitales. having the typical three-layered 

 integument (outer fleshy, middle stony, inner fleshy), the 

 apical beak of the nucellus, and the pollen chamber : but the 

 set of bundles which in cycads and cordaitales traverse the 

 outer fleshy layer are not present in Ginkgo, only the inner 

 bundles in the inner fleshy layer. A single mother- cell (very 

 rarely two mother-cells) functions, and lies among a mass 

 of tissue which may or not be sporogenous, representing a 

 many-layered archesporium, but at any rate functions as a 

 tapetum nutritive tissue. In the mature mother-cell the 

 reduced number of chromosomes is eight ; the four spores 

 formed by division of the mother-cell may be in a row, or 

 three with the upper cell divided longitndinally. The tapetal 

 tissue increases in bulk by division, during the tetred division 

 of the mother-cell, and actively encroaches upon the surround- 

 ing nucellar tissue ; the tapetal zone in its turn begins to be 

 destroyed by the encroachment of the young gametophyte 

 (endosperm), which invades and destroys the tapetal tissue 

 and the surrounding nucellar tissue. The megaspore. which 

 has a well-de\ eloped membrane, enlarges greatly, and its 

 nucleus lies at the micropyle end. 



Miss Carothers then describes the development of the 

 female prothallus (endosperm). The megaspore, after its 

 nucleus begins to divide, contains a large vacuole, hence the 

 nuclei formed by division are from the first arranged just 

 within the membrane, lying in the thin protoplasmic 

 lining layer. Up to the stage with sixty-four nuclei, 

 the free nuclear divisions are simultaneous, but later they 

 become irregular, until there are over two hundred and 

 fift}'-six nuclei ; meanwhile the whole ovule and the embryo- 

 sac are enlarging, and the megaspore membrane becoming 

 thicker. Previous to the formation of walls, a delicate but 

 distinct membrane appears on the outer surface of the 

 protoplasm of the embryo-sac, and to this membrane (not to 

 the megaspore membrane, as previously supposed) the first 

 walls of the endosperm are attached ; then by the nuclear 

 division and wall formation the prothallus becomes a solid 

 mass of tissue. Before this process is complete, the endosperm 

 becomes green, owing to formation of chlorophyll, as proved 

 by examination with the spectroscope, and the cells are soon 

 filled with starch grains — partly formed no doubt by the green 

 endosperm itself. 



The germin.ation of the pollen-grains, the formation of the 

 remarkable ciliated motile male cells, and the act of fertilisa- 

 tion of the two (sometimes three) archegonia were fully 

 described by Hirase in 1898, and little has been added to his 

 account. Various more recent writers have, however, worked 

 out the development of the embryo. The embryo of Ginkgo 

 was formerly regarded as peculiar among Gymnosperms, owing 

 to the absence of an elongated suspensor, such as occurs in 

 Cycads and Conifers. In Bcnncttitcs. however, there was 

 no suspensor. Arnoldi (1903) showed that the elongated 



pro-embryo of Ginkgo is differentiated into three regions : — 

 (1) a micropylar haustorial region (2) a middle suspensor 

 region, and (3) an apical region producing the embryo itself; 

 this brings Ginkgo into line with the typical embryology of 

 other Gymnosperms, though the suspensor remains short and 

 thick. Another peculiarity in the embryo of Ginkgo — the 

 complete filling of the fertilised egg with proembryonic tissue 

 — has lost significance by the discovery that the same thing 

 occurs in Dioon and probably some other Cycads. Lyon (1 904) 

 gave a very full account of the de\elopment of the embryo in 

 Ginkgo, greatly amplifying the descriptions given by Stras- 

 burger (1872) and other previous authors; in some cases 

 the pro-embryo apparently produces two embryos ; three 

 cotyledons occur not uncommonly, instead of two, 



CHEMISTRY, 



By C. .AiNswoRTH Mitchell. B.\. (Oxou.), F.l.C. 



USE OF COPPER FOR WATER PIPES.— The owners 

 of a house in Paris, being anxious to instal copper pipes 

 instead of the usual leaden pipes for their water supply, 

 applied to the Prefect of the Seine for the necessary per- 

 mission. The c|uestion was therefore submitted to the 

 Municipal Laboratory in Paris, which reported unfavourably 

 upon the proposal, and also to the Council of Public Hygiene, 

 the head of which. Professor .\rmand Gautier, spoke strongly 

 in its favour. In his report he pointed out that lead was 

 really an unsuitable metal for water pipes, and that the purer 

 the water the more rapidly was the metal attacked. .'Although 

 the presence of sulphates in most drinking waters rendered 

 the risk of poisoning insignificant, it was probable, he pointed 

 out, that there would be trouble in the immediate future from 

 this cause in the case of the particularly soft waters which 

 have recently been supplied to the towns in the West of 

 France. 



This danger would be entirely obviated by the use of copper 

 pipes, for although a dose of fifty to one hundred milligraunnes 

 of copper sulphate was poisonous, the system could soon 

 become acclimatised to much greater ijuantities, and, as a 

 matter of fact, several milligrammes were frequently absorbed 

 daily without any ill effect in the food which throughout 

 France was cooked in copper vessels. Experiments made by 

 M. Gautier upon himself had shown that no ill effects were 

 produced by acid food that had been cooked in copper utensils. 

 Long before the poisonous dose had been reached a liquid 

 contaminated with copper would have acquired a harsh 

 metallic taste which would put the consumer upon his guard, 

 and this would be the worst that could happen in the excep- 

 tional cases where water would attack the pipes. With leaden 

 pipes, however, a poisonous dose could be present in the water 

 without its being detected by the taste. There was, therefore, 

 in his opinion, only benefit to be deri\ ed from the substitution 

 of copper for leaden pipes for the supply of water to houses. 



The report of Professor Gautier. the chief points of which 

 are summarised above, is published in the current issue of 

 Biologica (1911, VI, 79). His conchisions agree with those 

 of the American chemists and doctors who have advocated 

 the use of copper sulphate and metallic copper for the 

 sterilisation of drinking water. Readers of these columns 

 will also remember that copper sulphate is now used upon a 

 large scale in America for destroying the pond scum upon the 

 water reservoirs. 



GRECIAN BLACK ENAMEL.— The rich black lustre 

 upon Grecian and Roman pottery has long been attributed to 

 the presence of manganese, but according to M. L. Franchet, 

 who has recently studied the question \Coniptcs Rcndiis, 

 1191, CLII, 1097), the proportion of manganese is altogether 

 insufficient to produce the effect upon the ancient vases. The 

 suggestion made by Verneuil that the (jreeks made use of 

 metallic iron obtained by the reduction of an oxide, is not 

 in accord with the statements of ancient writers, who assert 

 that natural products were employed. 



Since the mineral magnetite was known to the Romans, 

 who obtained it from Piedmont, M. Franchet has made 



