December ii, 1919] 



NATURE 



0/ o 



But at the point A, near the beginning of totality, 

 the left-hand ray, coming from a star about i° west of 

 the sun, would bo travelling from a denser to a rarer 

 medium ; the right-hand ray, coming from a star i° 

 east of the sun, from a rarer to a denser medium. 

 Hence both rays would be deflected towards the east, 

 and the distance between the stars would be unaffected. 

 The same thing would happen at C, the deflection of 

 both lieing now to the west. It is only at the middle 

 point B that the apparent distance could bo altered, 

 and here the temperature-gradient vanishes. 



.^^ 



B..- 



A.-'' 



Lenoth or MAjoD AXIS or SHADOW ellipse. 



H) fB U *r tC CO^fCt 



Moreover, at .\ and C there would be no tempera- 

 ture gradient (and therefore no optical shift) in the 

 plane perpendicular to the plane of the diagram. But 

 the measures showed equal shifts in all directions. In 

 actual fact, the presence of much cloud must have 

 made the temperature gradients irregular ; the fall 

 of temperature at Sobral during totalitv was verv 

 small, doubtless owing to the cloudiness before 

 totalitv. .\. C. D. Crommeltn. 



Sex-Phenomena in the Common Limpet 



(Patella vulgata). 



In the course of investigations on the rate of growth 

 and the age at which breeding begins the common 

 limpet was e.\amined, and the following interesting 

 phenomena were observed : — .\ preliminarv e.xamina- 

 tion of batches of limpets of about 2 cm. long, and later 

 still of smaller specimens, revealed 

 the occurrence of a large propor- 

 tion of males. The proportion of 

 males was so high irjdeed as to 

 give strong suspicions of a change 

 of sex from male to female (i.e. 

 protandric hermaphroditism), and 

 a sample of about icoo small ones 

 less than 1 in. was therefore col- 

 lected from cement piles belweeii 

 3 ft. and g ft. above low-water 

 springs at the Great Western 

 Docks, Plymouth, and the sex 

 examined and recorded. .\s the 

 common limpet has no penis or 

 uterus, it is nccessarv to examine 

 the internal sex-organ (the fjonad) 

 to determine the sex. Of the 

 1 102 limpets collected, 167 were 

 rejected as being found by experi- 

 ence to be too small to show- 

 development of the internal sex- 

 organ ; these were mainly about 

 13 mm. long. Of the remainder, 

 i6g (mostlv about 14 mm.) showed 

 no development of gonad on 

 examination, 64 were females 

 inostly about 2 cm. long, and the 

 remaining 702 were males mostly 



about 15 mm. to 20 mm. long. These males probably 

 compri.se the bulk of the limpets in this sample in 

 their first spawning year. 



This result confirmed the suspicion that all limpets 

 might be born as males, and to determine whether 



45 cm. was examined from the same locality. Of 

 these 255 were males, 3 of indeterminate sex, and 

 334 females. These figures indicated sex-change, but 

 were not sufficiently definite ; hence a further sample 

 of about KKK) very large limpets, from 5 cm. to 

 75 cm. in length, was obtained from Looe Island 

 and examined, with the result that 693 were found 

 to be females, 18 were of indeterminate sex, and 

 301 were males. Some of these latter males were very 

 big, ranging up to 65 cm. in length; soine males, 

 therefore, may live several years before changing intc 

 females, if, indeed, these larger males ever change 

 into females. At the same tiine as these very large 

 limpets were collected a batch of tiny limpets was 

 taken from the same localitv and the sex examined. 

 Of this sample of 1233 tiny limpets, 138 were rejected 

 as being too small (i.e. circa 13 mm. and less), and of 

 the remainder, all of which were examined, 944 were 

 males (mainly from 16 mm. to 20 mm.); 113 showed 

 no gonad developed, and were iriostly small, about 

 15 mm. in length; and the remainder, only 38, were 

 females, and mostly about 2 cm. long. 



.\mongst the 102 females recorded in the 2030 

 voung examined, however, 4 were found of a size 

 about 15 mm. long. The writer thinks that this small 

 proportion of small females will be found to be dwarf 

 females analogous to those described bv Conklin ' and 

 bv Orton " in the slipper Jimpets (Crepidula fornicata 

 and other species). In this investigation so far length 

 of shell has been taken as an indicator of age, and 

 doubtless on the average length is a good indicator 

 for a given locality. But the rate of growth is known 

 to be variable from as yet unpublished work, and it is 

 surmised that the smallest female limpets are two- 

 vear-old forms in which growth has been stunted, and 

 that such specimens are therefore older than their size 

 indicates. In this case it would appear that sex- 

 change may occur any time after the first spawning 

 season. One readv method of testing this view would 



TOTAL OF SAMPLES 



MILLIMETRE GROUPS 



25/ lSlit27V 



Length and se.v analysis of a sample of 1102 young limpets under i in. in length collected at 

 random at Plj mouth, October 31, igig. 



be to obtain a large number of young limpets of 

 known age, and this letter is written mainly with the 

 view of obtaining information of any dock-walls, slips, 



1 "The Embryology of Crepidula." By E. G. Conklin. Journal of 

 Morphology^ vol. xiii., 



p. '2, 1807. 

 rhn^- -.11 ^JinrKiP Intr, f».rml*.K n cMmnlf nf Tlviiit finn , ' ' On ' the Occurrence of Protandric Hermaphroditism in Crepidula 



tney all change mto lemaies a sample oi aoout ooo i /^rnicata." By j. H. Onon. Proc. Roy. .Soc, B, vol. ix.xxi., p. 480 

 larger but medium-sized specimens from 3 cm. to 1909. 



NO. 2615, VOL. 104] 



