October 31, 1912] 



NATURE 



!59 



although the pretty rosettes differ considerably in their 

 arrangement. Photographs of the fibrils detached 

 from torn valves are given. 



Mr. C. B. Crampton has followed up his recent 

 work on the vegetation of Caithness by a series of 

 articles in The Scottish Botanical Review, of which a 

 reprint has reached us. This important paper dis- 

 cusses in detail the geological relations of stable and 

 migratory plant-formations. After dealing with the 

 conceptions of the plant-formation given by various 

 ecolop-ical writers, the author distinguishes two classes 

 of habitats differing in the changes or successions 

 shown bv the vegetation and in the limits set to the 

 stability attainable by this vegetation. The two 

 classes of plant-formations, which tend to overlap and 

 invade each other's territory owing to the migratory 

 nature of the geological agents of surface change, 

 are : (i) stable formations, the plant-associations of 

 which have their centres of distribution on ground 

 which has for a long period been comparatively stable 

 from the geological point of view ; and (2) migratory 

 formations, the associations of which have their dis- 

 tribution centres in areas within the sphere of influ- 

 ence of the geological agents of surface change. The 

 author works out his conceptions in detail, applying 

 them successfully to the various types of vegetation 

 found in this country in particular, and gives a useful 

 bibliography of recent ecological literature. 



An interesting note on the cold August and Septem- 

 ber in London is published in Symons's Meteorological 

 Magazine for October. Dr. Mill states that the long 

 record at Camden Square (N.W. London), dating from 

 1858, contains no instance of any previous August or 

 September with a lower mean temperature. The 

 mean for August, 579°, was 4*4° below the average, 

 and it was the coolest August in the fifty-five years' 

 record. In September the mean was 54' 1°, or 3"6° 

 below the average. The shade maxima records are 

 more remarkable than the minima ; in August the 

 mean shade maximum was 66'6°. August, i860 and 

 1912, are the only months of that name in which the 

 shade temperature failed to reach 77°. In 1912 the 

 highest recorded was 73'2°. In September the mean 

 maximum, 62*4°, was the lowest on record for that 

 month, and the absolute maximum, 69'4°, was the 

 lowest recorded in any September. Dr. Mill remarks 

 that it is of considerable interest that fifteen consecu- 

 tive months with mean temperatures above the 

 average, May, 191 1, to July, 1912, should be followed 

 by two months of unprecedep.tedly low temperature. 



The meteorological charts of the Indian Ocean for 

 November, issued by the Meteorological Office and by 

 the U.S. Weather Bureau, both contain useful articles 

 on the cyclones of the South Indian Ocean. Both are 

 based to a great extent on the cyclone tracks compiled 

 by the late Dr. Meldrum, of Mauritius, and his suc- 

 cessor, published by the Meteorological Committee. 

 The cyclone season for this part of the ocean is from 

 November to May inclusive, but storms are occasion- 

 ally met with in other months; the maximum fre- 

 quency is from December to March. The majority 

 of the storms follow parabolic tracks, and, as a rule, 

 NO. 2244, VOL. go] 



recurve between latitude 20° and 22° S. Broadly 

 speaking, the storms are found to originate some- 

 where along the parallel of 10° S. ; " thence they travel 

 south-westward over a track that trends more and 

 more southerly until the vertex is attained ; afterwards 

 the track recurves to the south-eastward." The 

 Meteorological Office chart also gives very interesting 

 details respecting the behaviour of storms in the Bay 

 of Bengal and the Arabian Sea. 



Those who are interested in the studv of the geo- 

 metry of the triangle will welcome the publication of 

 a monograph of more than fifty pages bv the late Prof. 

 G. Sidler in the Mitteilitiigen der naturjorschenden 

 Gesellschaft in Bern for 191 1, published this year. 

 The original manuscript was completed in 1902, but 

 a difficulty occurred with the figures, which were of 

 great complexity, and the arrangements for publica- 

 tion were terminated by the death of the author. The 

 manuscript, which was in the possession of Prof. 

 Sidler's widow, has now been edited by Dr. O. 

 Schenker. 



A p-ivPER by Mr. A. Ferguson on the genesis of 

 logarithms, in Science Progress for July, should prove 

 interesting and useful reading to mathematical 

 teachers and others. Mr. Ferguson has carefully 

 studied Napier's " Mirifici Canonis Logarithmorum 

 Descriptio " and other authorities, and many of the 

 facts brought to light are little known. The paper 

 gives in a simple and intelligible form the methods 

 by which logarithmic tables were constructed from 

 first principles without the use of the modern infinite 

 series or the calculus, and it affords historical evidence 

 in support of the modern methods of teaching the use 

 of logarithms without assuming the definitions of 

 negative and fractional indices. 



As is well recognised, the old system of gauging 

 the strength of concentrated radium preparations in 

 terms of uranium, the "activity" being expressed as 

 so many million uranium units, is practically mean- 

 ingless, on account of the impossibility of comparing 

 together the radio-activity of two elements, so different 

 both in the character and in the intensity of the 

 radiations they emit. The announcement by Messrs. 

 Hopkins and Williams, Ltd., Hatton Garden, E.C., 

 that in future their radium preparations will be sold 

 on the basis of the actual quantities of radium they 

 contain is therefore a step in the right direction. We 

 understand that they measure their preparations 

 against a standard certified by Mme. Curie to contain 

 a definite quantity of radium bromide, and have for 

 disposal a considerable quantity containing from 90 to 

 92 per cent, of radium bromide. 



Since the echelon spectroscope first disclosed the 

 complex structure of many spectral lines previously 

 thought to be simple, the question of the exact com- 

 position of these lines has become a serious one owing 

 to the divergent results obtained by observers working 

 with different instruments. The green mercury line 

 5461 X 10-' cm., for example, has by various observers 

 been resolved into from seven fo twenty components. 

 The higher numbers appear from the more recent work 

 to be due to spurious lines produced by internal re- 



