LUNAR PERIODICITY IN LIVING ORGANISMS 279 



both the first and third quarters of the moon from October to 

 May (Hempelmann, Zoologica, 25, 191 1). The latter bilunar, 

 i.e. apparently tidal, periodicity is remarkable since the tidal 

 range at Naples is much less than at Woods Hole, yet swarming 

 at the last-mentioned place is not correlated with the tides. 

 The behaviour of Nereis is the more anomalous since twenty 

 years' observation in the Thames estuary by Sorby (Journ. 

 Linn. Soc, 29, 1906) failed to detect any lunar rhythm in the 

 swarming of this worm. Amphitrite, also a Polychsete, lays its 

 eggs at new and full moon spring tides at Woods Hole (Scott, 

 Biol. Bull., 17, 1909). Scott supposes that the higher tempera- 

 ture of the sand flats on which the animals live, due to the 

 abnormally low water at spring tides, together with a more 

 abundant food supply at these times, are the immediate causes 

 of the tidal reproductive cycle. 



Among Molluscs the only case I know of was communicated 

 to me in a letter by Dr. W. J. Crozier, who states that at Woods 

 Hole a chiton spawns at the time of full moon. 



To my knowledge the only other case of reproductive rhythm 

 in animals correlated with the lunar period is in the human 

 race. Arrhenius {Skand. Arch. /. Physiol., 8, 1898) showed by 

 a statistical investigation of 25,000 cases that there exists a low 

 correlation between the frequency of births and the tropical 

 lunar period of 27-32 days.^ Rather more births take place in 

 a certain part of the tropical lunar month than in the remainder. 

 Since this nativity rhythm might depend upon a periodicity in 

 menstruation, Arrhenius then set to work to find out whether 

 the latter exists. His material was furnished by 12,000 cases 

 from Stockholm maternity hospitals, where the date of the last 

 menstruation before the onset of pregnancy is recorded for each 

 patient admitted. The same kind of periodicity was found as 

 in the case of the births, but to a more pronounced degree. 

 Taking the day on which the moon passes through the ecliptic 

 from North to South as the first day of the tropical lunar month, 

 the curve giving the frequencies of menstruation for each day in 

 the month gradually descends to a minimum on the eighteenth 

 day, when there are 6' 5 per cent, fewer cases than the average 

 for all the days of the month, and then gradually rises again to 

 reach a maximum at another part of the month. In each 

 individual the onset of menstruation depends, of course, upon 

 a number of physiological and psychical causes, which occur 

 without any rhythm. In addition there is a rhythmic causative 

 factor having the period of the revolution of the moon. This 



^ This is the time taken by the moon to make one circuit of the heavens. 

 The synodical month or lunation of 29* 53 days is the time in which the moon 

 goes through her phases. 



