394 A. W. Evans — Hmcaiian Hepaticoe of the Tribe Juhuloidece. 



In the genus Juhula^ which is intermediate in some respects 

 between F)'ullania and the Lejeuneeae, the plants are distinctly 

 green and are never tinged with red. The leaf -lobe is attached to 

 the axis by a long oblique line of insertion, the lobule is distant 

 from the axis and is therefore not attached to it at all, and the 

 stylus, which is reduced to a single cell, becomes obsolete very early : 

 it appears, therefore, as if no part of the leaf, except the lobe itself, 

 were attached to the axis. The lobes are often dentate or ciliate. 

 The lobe beneath which a branch is situated is more oblique than 

 the others and is attached partly to the axis and partly to the branch. 

 The first underleaf of a branch is usually a simple lanceolate lamina : 

 it is much pushed out of position, being attached partly to the axis 

 and partly to the branch, and its line of insertion does not meet that 

 of the lobe, which is here obviously without a lobule. Innovations, 

 like those of the Lejeuneete, are present ; in typical cases, there are 

 two innovations for each inflorescence, though it is not unusual to 

 find only one of them developed. In Juhula Hutchinsice,* the male 

 branches apparently arise in the same way as those of the following 

 subtribe, being boi'ne behind small but otherwise normal and lobu- 

 late leaves : this peculiai'ity, however, is not constant for the genus 

 and is not found in the Hawaiian species. The stalk of the capsule 

 is formed of only two concentric layers of cells, as in the Lejeuneeae. 



1. FRULLANIA Eadcli. 



In the last published list of Hawaiian Hepatic8e,f ten species of 

 Frullania are enumerated. An eleventh species, F. Oahuensis, first 

 collected b}^ Meyen, is not included here. Although published in 

 1843, this species was omitted from the Synopsis Hepaticarum, per- 

 haps through an oversight, and has not since appeared in hepatico- 

 logical literature. Several of these eleven species cannot be main- 

 tained. Three of them, F. arietina, F, Kunzei, and F. sqiiarrosa, 

 are listed on incori-ect determinations, and three of the others, 

 F. explicata, F. oceanica, and F. Helleri, are synonyms. Ang- 

 strom's !>. Sandvicensis, on the other hand, is a mixed species and 

 was based on two pei-fectly distinct plants. The six species which I 

 have been able to distinguish fall naturally into three of Spruce's 

 subgenera and may be identified as follows : 



* Cf. Leitgeb, Unters. iiber die Leberm. ii, 37. 1875. 



f Stephani, Hepaticae sandvicenses. Bull, de I'Herb. Boissier, v, 842. 1897. 



