82 
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES 
Series 4, Volume 65, 28 Sept. 2018, No. 2 
the 1980s. Although most of these are no longer extant, some still remain as street trees in that city. 
Based on these data, this species is accorded a preliminary conservation assessment of Least Con¬ 
cern (LC). 
Discussion. — In the protologue, Watson (1888) provided the collection information cited 
above. The holotype has been cited by various authors (e.g., Richardson 1972; Daniel 1995) as 
being at US. Indeed, a specimen there (accession: 1320518 with image number: 00136468) was 
annotated by D. Gibson and T. Daniel as the holotype. This somewhat fragmentary specimen, orig¬ 
inally in H. Rusby’s herbarium, bears von Tiirckheim’s number 856 (as indicated in the proto¬ 
logue), but notes a collection date of February 1886. The date of collection of the specimen from 
Pansamala indicated in the protologue is May 1887. The specimen at GH, where Watson conduct¬ 
ed his studies, bears the date noted in the protologue and an annotation written by Watson stating, 
“Louteridium Donnell-Smithii, Watson-n. gen.” Therefore the specimen at GH indicated above 
conforms to the information in the protologue. The earlier annotations by Gibson and Daniel of the 
specimen at US are considered to be erroneous, and the specimen at GH noted above is treated as 
the holotype. This flowering and fruiting specimen lacks leaves, which are described in the proto¬ 
logue. However, drawings of the leaves are on another sheet at GH bearing the same barcode num¬ 
ber. If the drawings are original material, the holotype would be considered as mounted on two 
sheets, and if they were made subsequently, the sheet with an inflorescence, flowers, and fruits 
would be considered as the holotype. Other collections that bear von Turckheim’s number 856, but 
with a date of February 1886, are located at the following herbaria: K, NY, P, PH, US (although 
one of three specimens at US has both “1887” and February 1886 on it). Whether these represent 
isotypes with an incorrect date (i.e., actually collected in May 1887) or the same collection 
number was used for gatherings of this species during different visits to Pansamala on various dates 
remains unknown. 
Daniel (2010) noted variation in the timing of anthesis, color of corollas, and diurnal visitation 
to dark-colored flowers by hummingbirds in this species. There is a possible correlation with cream 
to green-yellow corollas at lower elevations and chestnut to maroon corollas at higher elevations. 
Based on specimens of L. donnell-smithii at F, which are mostly from Guatemala, plants with 
cream to green-yellow corollas occur from 75 to 630 m, whereas those with chestnut to maroon 
corollas occur at 1000 to 2100 m. Figures 4C and D show these extremes of corolla coloration in 
Guatemala. Other flowers show various degrees of maroon coloration on cream-yellowish corol¬ 
las. Based on a larger sample of herbarium specimens that include data on corolla color and eleva¬ 
tion, it would appear that both the cream to green-yellow and chestnut to maroon color forms occur 
in Chiapas, Belize, and Guatemala, but that there is no clear difference in elevational ranges of 
these forms in Chiapas and Belize. However, it is somewhat difficult to correlate corolla color and 
elevation from herbarium specimens because of variation in the terminology of colors used by 
collectors. Additional studies that address the different (or at least partially so) flowering times of 
the different color forms and their primary pollinators, as well as any correlation of floral color with 
elevation, are desirable. 
The inflorescence of L. donnell-smithii is commonly a somewhat contracted thyrse at the prox¬ 
imal nodes (with dichasial peduncles usually to 5 mm long) and a raceme (with sessile dichasia) 
distally. Several collections (e.g., Armour & Chable 6047, Hawkins 981, Vargas et al. 1183 ) have 
exceptionally elongate peduncles (to 90 mm long) with expanded dichasia (e.g., secondary pedun¬ 
cles to 55 mm long). These latter dichasia are usually only at the proximal 1 to 3 nodes. 
Additional specimens examined.— MEXICO. Chiapas: Las Nubes, F. Billiet 8889 (BR); 
Mpio. Ocosingo, W end of Laguna Ocotal Grande, [ca. 16°50’26.15”N, 091°28’5.95”W], 
D. Breedlove 15709 (DS, F, LL, MICH, NY, US); Mpio. Las Margaritas, western side of Laguna 
