174 
PACIFIC SCIENCE, VoL XIII, April, 1959 
most of them entire. Four of them have some 
of the blades slightly sinuate, and one has a 
single large leaf strongly sinuate. From this 
review of the characters of the leaves, the 
writer is convinced that the shapes and shal- 
low lobing of the leaves mentioned are mere 
fluctuations in a single population and usually 
on the same individual branch. Hence, the 
specific and varietal names based upon plants 
showing these characters are here reduced to 
the synonymy of S. Nelsoni Dunal. 
Early there was described another variety, 
5. Nelsoni var. thomasiaefolium Seem. (Jour. 
Bot., Brit, and Foreign 1: 209, 1863), based 
upon a collection by T. Nuttall on Atoi 
( = Kauai) . This may well be the greatest ex- 
treme among the lobed-leaved plants. At 
hand is a photo of the type in the British 
Museum. Three years after publishing the 
variety Seeman again listed it (1866: 174) and 
repeated this original account; "foliis cordato- 
ovatis sinuato-lobatis, lobis (5-7) obtusis vel 
cordatis integris. . . . This has quite the look 
of Thomasia solanacea. Gay, and would prob- 
ably be described as a new species by anyone 
not having seen the evident transition there 
is in some specimens of what Nuttall has 
called 5. rotundi folium and A. Gray justly con- 
siders identical with the original S. Nelsoni, 
Dun., preserved at the British Museum. In 
these specimens some of the leaves have a 
tendency to become sinuato-lobate, whilst 
again several leaves of my var. thomasiaefolium 
are cordate and entire.” The photo of the 
holotype confirms this, that there are several 
leaves that are cordate and entire, but mostly 
they are strongly sinuately lobed, with two 
large, sinuate lobes on a side, cut half-way to 
the midrib. Seeman well states the inconstancy 
in this group of the characters of leaf shape 
and lobing, and the writer has stated his ob- 
servations on the same points. However, no 
recent collections or any others seen show 
blades with the deep rounded lobes like those 
of var. thomasiaefolium. It is a marked extreme, 
and no intermediates have been seen that fill 
in the gap. Hence, until the variety is again 
collected and more knowledge can be gained 
of its variability and occurrence, it seems best 
to accept var. thomasiaefolium as a variety, pe- 
culiar to Kauai. The writer and his several 
assistants searched for it in December, 1947, 
on the sandy shores of Kauai without success. 
The treatment of this group by Dr. F. B. H. 
Brown (1931: 37) raises a question of typifica- 
tion. He gives a generalized description for 
5. Nelsoni Dunal (as nelsoni), then a detailed 
one for each of the four varieties, including 
var. typicum F. Br., new var. As var. typicum he 
recognized the plants with cordate, entire 
leaves. For this he cites only two collections, 
both in the Bishop Museum, and both from 
Moomomi, Molokai: J. F. Rock, March 1910, 
sheet A (which lettering was added to Rock’s 
label in a later hand, apparently Brown’s); 
and C. N. Forbes no. 613. Mo. No type was 
designated in the publication, but in the 
Bishop Museum the specimens are marked: 
the Rock sheet A is marked in Brown’s writing, 
"Type of Descript.”; the Forbes 61 3. Mo. is 
marked in Brown’s writing, "Type! of amplif. 
descript.”; and there is a third, unlisted, 
sheet, Forbes 604. Mo., marked in Brown’s 
writing, "Type of amplif. descript.” On 
checking Forbes’s field number book, it is 
evident that both of his numbers written by 
him on his labels were wrong, that his collec- 
tion of this Solanum from the sand dunes at 
Moomomi was no. 607. Mo., and that both of 
his sheets should be corrected to so read, as 
does an unmounted duplicate. Thus Brown, 
in describing S. Nelsoni Dunal var. typica F. 
Br., published no choice of a type in the 
bulletin, which was issued by Christophersen 
and Caum, but in the herbarium he marked as 
type three sheets collected in 1910 and 1915 
by two different collectors. Not every possible 
detail concerning the selection of types is 
covered in the International Code of Botanical 
Nomenclature, but the practice is established 
and legalized. It is clearly improper to choose 
either the Forbes or the Rock collections to 
be the type of var. typica (now called var. 
Nelsoni) . The type must be the same specimen 
