October 21, 1922] 



NA TURE 



559 



the Persian Achsmenidae was divided between the 

 Parthians, capitalled at Ctesiphon, and the Kushan 

 dynasty of the Yue-chi, capitalled at Peshawar. 

 These four northern frontiers, Roman, Parthian, 

 Kushan, and Chinese, were consecutive, forming an 

 unbroken line from the mouth of the Rhine near 

 the modern Katwyk in Holland, 52 N., to the east 

 coast of Korea in about 41° N. South of the line 

 a vast array of established cities stretched for seven 

 thousand miles across Eurasia, in some parts protected 

 by natural barriers, in others defended by lines of 

 masonry fortification. North of the line were the 

 tents of nomads, huts of forest dwellers, and stockaded 

 defences of earth and wood. In the northern part 

 of modern Germany there were territories north of 

 the line which the Romans had abandoned as unten- 

 able or unprofitable. South of the line in Eastern 

 Europe was the district of Dacia which Augustus 

 preferred not to touch, but Trajan was compelled 

 to occupy. In this country the native people had 

 in the interval begun to construct masonry fortifi- 

 cations. 



In the course of an investigation of the geography 

 of capital cities, it was found that this northern 

 frontier of ancient cities, on the eve of the barbarian 

 irruption, has, within narrow limits of variation, the 

 same average temperature throughout. It is a true 

 annual isotherm, not an is< itherm reduced to sea 

 level. Along the European part is a line of modern 

 cities with meteorological observatories. The annual 

 temperatures of eight of these, strung out along the 

 length, has an average of 48°-6 F. Asia is not well 

 off for meteorological records near the line on the 

 south, and the second table consists of a list of 

 towns mostly under Russian rule just north of the 

 line where proper records have been kept. It will 

 be observed, therefore, that their temperatures are 

 rather lower than that along the frontier of the 

 ancient cities. The average temperature of these 

 eight towns north of the line is 47°'4 F. A very 

 long gap in these towns occurs between Kuldja and 

 Mukden, but the record for the Lukchun depression 

 in Chinese Turkestan, a little south of the frontier 

 yields a not inconsistent figure, if corrected for the 

 general height of the surrounding country, and that 

 of Peking is not discordant. Further east the 

 generalised isotherm of 48°-5 F. reaches the eastern 

 coast of Korea in about 41 ° N. (somewhat north of 



the peninsula portion of the country) which cannot 

 be very far from the frontier of its ancient cities. 



In the detached Roman possession of Britain the 

 inner and principal line of fortification had its western 

 terminal at Carlisle, where the temperature is 47°-8 F. 

 Eastwards of the continent of Eurasia the conquest 

 of the Japanese islands by their present masters 

 was only completed at a much later date than that 

 under consideration, but the Japanese derived their 

 culture from ancient China (mainly through Korea) 

 and it may therefore be significant that thev were 

 content to conquer, without colonising, Yezo, the 

 northern island, and that what is reckoned by the 

 Japanese as Japan proper, and is called by them 

 " Old Japan " does not include Yezo but stops 

 short with Honshiu, the mainland, and that the 

 annual isotherm of 48°-5 F. traverses the strait of 

 Tsugaru which separates Old Japan from Yezo. 



The fact that the annual temperature along this 

 immense line only varies within remarkably narrow 

 limits cannot be reasonably contested. If it be the 

 case that desiccation has occurred generally in Asia 

 along this line since the second century of our era 

 its probable effect would be to lessen the winter 

 and raise the summer temperature, leaving the 

 annual temperature much the same. 



The coincidence of frontier and true isotherm is 

 not a mere consequence of east and west barriers 

 of mountains, inland seas, and rivers, for these had 

 to be supplemented by long lines of fortification. 

 Neither was it due to unsuitabilitv of the southern 

 country to pastoral peoples, for in Asia there was 

 much coveted grazing land south of the settled 

 frontier. Precisely how far this coincidence is sig- 

 nificant it is yet difficult to say. 



The Mechanism of the Cochlea. 



MOST medical students have probably felt that 

 current physiological teaching provided them 

 with only a hazy conception of the mechanism for 

 hearing in the cochlea. Helmholtz put forward the 

 view that this organ contained a series of resonators, 

 which were differentiated like a set of piano strings, 

 so that each string vibrated only in response to one 

 particular note. It will be remembered that the 

 cochlea forms a spiral, which when unwound consists 

 of two chambers, placed one above the other, and 

 separated by the basilar membrane. At one end 

 (the base) of the cochlea, in the wall of the upper 

 chamber, is the window which is set in vibration by 

 the middle ear, while in the wall of the lower chamber 

 is a similar window whose function is to prevent the 

 pressure from changing inside the cochlea when the 

 upper window moves. Both chambers contain fluid, 

 and, at the other end (the apex) of the cochlea, the 

 chambers unite, for the basilar membrane ceases just 

 short of the apex. 



The suggestion that the fibres of the basilar 

 membrane can act as a resonating 'system has been 



NO. 2764, VOL. I IO] 



current since it was pointed out that their length 

 (measured across the canal) varied continuously from 

 the base to the apex. Now the fibres of a resonating 

 system must obey the laws which govern vibrating 

 strings, so that n, the number of vibrations of a 



string per sec, =~)\/~. where / is the length of a 



fibre, t is the tension, and m is the mass per unit 

 length. Gray showed in 1900 that the tension of 

 the fibres of the basilar membrane also varied from 

 the base to the apex, for while the spiral ligament 

 which attached the membrane to the outer wall of 

 the cochlea was very dense near the base, it was, 

 on the contrary, very slender near the apex. We 

 know, therefore, that the fibres of the basilar 

 membrane are differentiated for tension and length, 

 so that the short fibres near the base are under high 

 tension, and the long ones near the apex are under 

 low tension. To complete the requirements of the 

 formula for vibrating strings, it is only necessary to 

 discover a system by which the fibres are differentiated 

 for mass, which differentiation must, as the formula 



