December 15, 1904] 



NATURE 



159 



Father Cortie states that the widening of some oxygen 

 Jines in sun-spot spectra, particularly in the a. band, seems 

 to be a real phenomenon. 



Eclipse Observations. — Vol. iii. of the Annalen of the 

 Royal University Observatory of Strassburg, edited by Dr. 

 E. Becker, the director, contains the results of the helio- 

 nieter observations of the total solar eclipse of May 28, 

 1900, and of the lunar eclipses which took place on 

 January 28, 1888, May 11, 1902 (partial eclipse), and April 

 ir, 1903, respectively. 



In the first part Prof. Kobold gives the results of a 

 number of observations made in order to determine the 

 reduction elements of the heliometer, and then applies them 

 to the observational results obtained during the solar 

 «clipse of 1900. Finally, he gives the corrections to the 

 .previously determined positions. In part ii. the same 

 observer discusses the observations of the 1888 and 1892 

 ■eclipses of the moon, and gives the values obtained for the 

 radius of the earth's shadow, &c., finally comparing them 

 with the calculated values. 



In the third part HeiT C. W. Wirtz discusses the observ- 

 ations of the lunar eclipse of April 11, 1903, including the 

 corrections to the moon's place, the figure and size of the 

 ■earth's shadow, and the variations of the diameter of the 

 •crater Linn^ during the eclipse. The curve on which are 

 plotted the values of the last named quantity shows a con- 

 siderable increase in the diameter during the approach of 

 ■the earth's shadow to the crater, the maximum value 

 evidently occurring during the actual eclipse of Linn^. 



The Appe.\r.anxe of Sp.ark Lines in Arc Spectr.h. — An 

 interesting discussion of the conditions which lead to the 

 appearance of " spark " lines in arc spectra is published 

 in No. 4, vol. XX., of the Astrophysical Journal by Dr. 

 Henry Crew, of the North-western University, 111. Dr. 

 Crew made a number of experiments in which the Mg 

 line at \ 4481 appeared in the arc spectrum, and examined 

 the arc, simultaneously, with a Rowland grating spectro- 

 graph and a Duddell high-frequency oscillograph. 



The various conditions under which the arc was produced 

 ■were as follow : — (i and 2) current with negligible and 

 with large amount of inductance respectively ; (3) arc 

 fcroken by air blast ; (4) arc in atmosphere of coal gas. 



The reproductions of the oscillograph curves show the 

 current conditions during each experiment, and from a dis- 

 cussion of the results Dr. Crew arrives at the following 

 ■conclusions: — (i) A rapidly changing, high E.M.F. is a 

 probable conditio sine qua non for the appearance of spark 

 dines in arc spectra. (2) The effect of hydrogen and other 

 atmospheres in introducing spark lines is explained by the 

 fact that these atmospheres produce a more rapid break, 

 and this, in turn, introduces an extra E.M.F., which in 

 ■some way, as yet unknown, is responsible for the radiation 

 ■of the spark line. A possible explanation of the stellar 

 ■conditions which produce spark lines in the spectra of stars 

 is also discussed. 



The Royal Astronomical Society of Canada. — Founded 

 as the Astronomical and Physical Society of Toronto, the 

 name of this societv was changed in 1900 to that of the 

 Toronto Astronomical Society. In 1903 it was decided to 

 change its name to the Astronomical Society of Canada, 

 and in response to a petition the privilege of prefixing the 

 word " Royal " to its name was granted, so that the full 

 title of the society is now the above heading. We hope 

 that this now national society will be a stimulus to the 

 promotion and diffusion of astronomical science, and that 

 its influence will be greatly extended. We have before us 

 the volume containing the selected papers and proceedings 

 for the years 1902 and 1903, edited by A. Harvey; the 

 varied topics there dealt with bid fair for the future of the 

 society. Among some of the papers may be mentioned the 

 address of the president, R. F. Stupart, director of the 

 Magnetic and Meteorological Observatory of Toronto, in 

 -which is an account of the history and work of the institu- 

 tion. W. H. S. Monck gives a catalogue of aerolites, 

 arranged in order of the months in which they fell. There 

 is a brief account of the present astronomical equipment of 

 Canada as a whole, and a discussion on papers dealing with 

 ■solar phenomena and terrestrial effects. The volume con- 

 cludes with an account of women's work in astronomy, by 

 Miss E. A. Dent. 



THE FIRST TRUE MAPS. 

 T N the history of cartography, in the development of maps 

 and map-making, there is perhaps nothing quite com- 

 parable to the first appearance of the " portolani " or 

 " handy charts " at the close of the thirteenth and the 

 beginning of the fourteenth century. For the portolani, 

 the first true sea-charts, are also the first true maps of 

 any kind — the earliest designs in which any part of the 

 earth-surface is laid down from actual observation of close 

 and continuous character. 



By the term " portolani " we intend, of course, to refer 

 to that great series of coast-plans of which the earliest 

 known examples belong to the first decade of the fourteenth 

 century (a.d. 1300-13 10); which are traceable to a very 

 few, perhaps to two or three (now lost), originals; which 

 may be extended to cover at least 500 designs (reaching 

 down to the end of the sixteenth century) ; and were 

 primarily intended to serve as practical guides to mariners 

 and merchants in the seaports of the Mediterranean and 

 Black Sea. 



These plans of practical navigators — of men whose liveli- 

 hood largely depended on their knowledge of nature and 

 their close observation of natural features — are a remark- 

 able contrast, in their almost modern accuracy, to the 

 results of the older literary or theological geography as we 

 have them in the Hereford or Ebstorf maps (both of the 

 very same period as the oldest existing portolans, c. a.d. 

 1300). They have never yet received adequate attention 

 from English geographers (as from Nordenskjbld the 

 Swede, Fischer the German, or Uzielli the Italian), and 

 the problem of their sudden appearance in such comparative 

 perfection is surely deserving of more study, and capable 

 of fuller e.xplanation, than it has yet received. Certain 

 assumptions may perhaps be made without danger. The 

 portolano type was not the invention of one man, of one 

 year, of one decade. It did not spring from any school or 

 any example of mediaeval student-map. It was the final 

 result of centuries' experience — the outcome of the notes, 

 plans, and oral tradition of generations of pilots and 

 captains. Skipper-charts of certain important and much- 

 frequented sections of the coast trade-routes were probably 

 combined, by slow degrees, into a coast-chart of the 

 Mediterranean basin as a whole. It may be that the 

 sketches of small portions of shore-line which we have in 

 fifteenth century manuscripts of Leonardo Dati's poem 

 " La Sfera " are really copies, but slightly modified, of 

 such old skipper-charts — reaching back, perhaps, to the 

 eleventh century, and forming the very earliest indications 

 of that new scientific geography in which the compass 

 played so great a part. If this surmise is correct, the open- 

 ing of the medieval Renaissance, in the generations immedi- 

 ately preceding the Crusades, was accompanied by the 

 oldest embryonic forms of modern cartography. 



Once more, it may be that the sea-chart which is 

 mentioned in connection with the Seventh Crusade (of 

 A.D. 1270), and which St. Louis apparently employed to aid 

 his attack on Tunis, was a portolan, or a sectional chart 

 of the North African coast of portolan type. It may be 

 that the charta noticed in Raymond LuUi's " Arbor 

 Scientiae " (about a.d. 1300) as necessary for sailors — along 

 with the compass, needle, and " star of the sea " — was a 

 work of the same kind. It may be that Andrea Blanco's 

 planisphere of 1436 is a re-edition of a " handy-map " of 

 the thirteenth century. But the oldest certain examples of 

 the type we are concerned with, which have been discovered 

 up to the present, are the Carte pisane and the first design 

 of Giovanni de Carignano, both belonging to the opening 

 years of the fourteenth century, while the oldest dated 

 portolan is the first of Pietro Vesconte (or Visconti), 

 executed in 1311. 



And when, with these and the next few examples, we 

 get at last our full coast-chart of the Mediterranean basin, 

 what is its character? 



It is a map without graduation, embracing only the coast 

 lines and the towns and natural features in the immediate 

 neighbourhood of the coast. But though it is restricted, it 

 has extraordinary merits in its own field. Its delineation 

 of the shores of the Mare Internum, from the Straits of 

 Gibraltar to the extreme east of the Black Sea, is markedly 

 superior to anything of earlier date — even to the Madaba 



NO. 1833, '^'OL. 71] 



