198 PERCY SLADEN TRUST EXPEDITION 



a series of oval and ovoid forms, on to specimens with heart-shaped and triangular gonads, 

 providing that the specimens are sufficiently numerous and taken over a large area. This 

 method of determination leads to a considerable reduction in the number of described 

 species, and practically it means that the genus should have only a single species. 



Another character which has been used as an aid to the determination of species is 

 the number of centripetal canals. These increase in number by age and by the growth 

 of the umbrella, so that taken alone they are not very reliable. Other characters have 

 also been used, such as the length of the peduncle and the shape of the umbrella. By 

 taking each character separately one can connect together specimens which have every 

 appearance of being quite distinct species. 



Some of the species have no doubt been described from single specimens, every 

 character has been used for the purpose, especially the shape of the gonads and no 

 allowance made for variation or growth. Apparently a description and figure based upon 

 a single or just a few individuals are not sufficient in the case of Liriope. It has, however, 

 occurred to me that by taking a sufficient number of adults at the same place and time, 

 and using all the possible characters, one would be able to obtain an impression of the 

 type of that particular lot of specimens, and produce a sketch not of an individual but 

 of the characters of the type. By adopting this method I think there is a chance of 

 isolating species or at all events local races. 



The number of specimens in the " Sealark " collection was not sufficient for finding 

 out the exact number of local races or species living within the area covered by the voyage. 

 After rejecting early stages and bad specimens the number left for the purpose was small. 

 I am able, however, to isolate two races or species, and got on to the track of a third one. 



35. Liriope tetraphylla (Chamisso et Eysenhardt) 1821. 



The specimens with triangular gonads I have placed under the old name of Liriope 

 tetraphylla, and they are very similar to Vanhoffen's figure of Liriope tetraphylla (1902, 



Taf 10, fig. 14). 



The umbrella is thin, 5 — 8 mm. in diameter. The length of the peduncle is about 

 twice the diameter of the umbrella. Velum very broad. In each quadrant there are three 

 centripetal canals, the interradial the longest extending up to or nearly to the top of the 

 gonads, the adradial canals about half the length of the interradial. The gonads are of 

 the triangular type, about equilateral, either with angular or rounded corners and do not 

 extend down to the margin of the umbrella. 



36. Liriope, sp. 1 



The second species or race I leave without a name, as I have not been able to find 

 a published figure showing the characters of the race, and the condition of the specimens 

 is not suitable for drawing. 



The umbrella is thin, about 5 mm. in diameter. The peduncle is a little longer than 

 the diameter of the umbrella. Only one centripetal canal (interradial) in each quadrant. 

 The gonads are longer than they are broad, either oval or elongated shield-shaped. 



The great differences between the two species are in the shape of the gonads and 



