ON 4 6 OA Set - < by " ahh tt tnt @ a0 ee Aae J =e’ On he tee het 4 -® 9044 a* * y pee Se Hee 5) ges > 4. . . 6 ae s ' + ue 4 14: s ov are e14- Fa 6 x ities : : ‘ ; i ; r er pots bi Bs Rees See vA Pups © y ‘ es “ : Bate eb. 4@@,4 e444 3 O44 s ” 2 SPAT er ey eee Aste hg ies : A, TOL, : AS 8 , ‘ be Yo: Kan 10m 4 tot BF RSA yakeo ple ah : . te’ isis heer ae art } i! sate ie ts ait 7 tari beatet Bene: : seth Ls dle t? Ee 3) Oe pea tne 4 es baw ae ; Ase. * ne any 4 rau ‘ i" ¢ bs “pe € th Jey Sepeby t cheat fas ote ' ts Bika: teh erat = ¥ -#e tah $34 ; Jon 4) 0 : ey ery - “Bi ae =e a eee a» +> a bayee 1 ‘' = gate * if Merl tee pide oe Be teketecs vr oH a ERD ae RUNS ae Meey wpler ray) € as thant ear Pe tle 48 Chas a’ “ae + a ae 17 ; > T iA qi getet ¢. re ; seat 4 toe eet Bae atid ae 4 + eRY i bi oto gine ie i # 3 “4 Cy ai4 dog, eres cole btrdia e's k') DETAIL peta et yh : + at? eed |e ta var bias tar A da oS As peopel oF S164 1828 289 de RS ees ae NC Le | ah.t. 452 Usa ie f AB a4 Fe ee het he i ; i f ‘ erg bari Fh J ‘4 - ' Ria fbn te dt Ado OL bb PRG s Toe 438 ‘te shih detey betes 14 rm ase MOG ot oes Pie WAS BAe 45 tr a YI, : +» its hé id fie Ae i¢ ern) he ’ J RS bn ig .¢ “ Sey Sma Stine eae 3 1h 89.47 ba O88 $154. dg lei tayte 5 ii eae tt apd) gts owe Ppl evens sant y og 9 OO La te boa ae Shed Oe et eet ed A Meals arstintart, A ATK Pe Uy, RO Oe amet 2) ee ab Peete Reread Wer PE fy Dx Be DY TMC DONEC OT OA LON DA A ALD Stall Aire td ity, Bibewed Arh Gi ge : Pibrtse ee prid Peart a pp der er it a © ‘ ; ' i" t a ht ’ ' eye) by LW MERIT TES] i rad Gea. : y ppeeeiercrere try. ” , ant Whe Wits nad ’ APA}: NS) ( : m4, a j 4 yaa 3 tri . spe 1?) uss ey Fe : _ ; ae ; ale dod i ivete: & rt) Y Sis 3 sce . yl = a ft @ 4 4: “ ste ‘ i, >) b+ 4 oe ~~, ere bin ts A ea eaiitt uaa “" a4, See A Pet 4 49,4 2%. 1.8 + FA a ee ye Pa Orpen WG AA te mS Ngh teat if Lis st ae pede LA Lae tye hat C4 af aide pet fib 4 * Od G:)4 A Roe a ea peak a ararin bed gid ae bas og #6 4.419 F iy ed Seg ea eta hg Cala 04 an deeds wd Aatcae | ORI LO Birt CK ONC CCRC AR BAUD Ph CNL a Pay ee Smee 88.4 Aaa ~ eae iz my ‘ wily } ve. ee bins age , iy iAia yn ~ » z , a by a > p ‘ oi 4 ae 5 ee, : xs " : 4 hua 4 é ‘ore ans Ft ae “ rf : e . ‘ Lit. ; > yaa, Rr teh tay” ae peta, A AN ay by, \ ie Ive, bier eee Pye ia ak She t An . . 7 ey) be 2 ¥ bb . % ‘ . ad ie 4 se - P. ok 1 ne Liss ae ee ae ‘g iw at Mov" RAR ey Ey, iy ae : eat Bybee 2 > a » > : ah ~ . as ane as ty ie: es 1 beh Oe * ~ < oy As Ee Yegh Bite G2 © ~~ ~ fs AN ; a. oh ba ST ee PC eee) eee PSST PST MEMOIRS OF THE NEw YORK BOTANICAL GARDEN VOLUME 8 ee I ECIMEOOE. ok, 2 ee) a W. H. Camp Plants collected in Ecuador by W. H. Camp. Lyman B. Smith, Jestis M. Idrobo, Bernice G. Schubert, A. C. Smith The botany of the Guayana Highland. Bassett Maguire, Richard S. Cowan, and John J. Wurdack, and Collaborators Vegetation of Nyasaland. Report on the Vernay Nyasaland Expedition ya SER sae DPSS, 7 Ott et on L. J. Brass Plants collected by the Vernay Nyasaland Expedition of 1946. J. P. M. Brenan and Collaborators . A taxonomic revision of the genus Macrolobium (Leguminosae- En SS a i eae ee Richard §. Cowan A revision of the genus Brachyotum (Tibouchineae-Melastomaceae). : John J. Wurdack Plants collected by the Vernay Nyasaland Expedition of 1946 ee ee eee J. P. M. Brenan and Collaborators Printed by The Science Press Lancaster, Pa. 25 87 161 191 257 343 409 _Dates of Issue of Volume 8 5 pp. LEON AS Fey > Ag 1932 2 pps RR EES Or 14 My 1953 29. pel Bits 2) ag an ee 2 ts Pe ee nO yess 16 O- L953 .5, pp. 409-510. 27F 1954 COPY 2 MEMOIRS 9 Eipy ies BOTA 1 IGA THE NEW YORK BOTANICAL 'GARGas ; GE & » . cf A) pce 0/0800 _inge Rio x e Rou a oy v @ Py popex §,000 0 Ky S et OO A v ~ MSF of = re py. eg ; > * LZAPOTE AREA iene a? ON, . ¢ S550 ee e Y A ¥ INDANZAS 3250 * hind Legend :— e ee e province boundaries ttt FALILOAR ———-vehicular road, quality variable —— —horse trails,usually passable ee route of party in Cord. Cutucu ieee X 6800 alt.in feet, barometric det. =e Agucar Scale : —a Rees GR ge ae 2 es aay kilometers 1952] - PLANT HUNTING IN ECUADOR AM BI / | nanngas R: Chano / 7900 : “a -Chy abe We Mes Q. Ning? i = © R ker, ig a =P Le / ; = ~ 2 eusd J */ | BASE & | > x. | CAMP | Ss 7 43450 = MENDEZ | s 4 a 1750/0 \ ese 6500 : a ~ Ae 2 ime ° zo 1400 © Y dle x 2 & p NS. \A s * ae @ 3850 @ LA-1150 ; =) ; SW SO ee ; : % | ee, pnalZ6 a” % a pyane” os Ds a te . a Bes 7 yy S <2 DA . x a 6, Se GSS Va YUQUIANZA NI 3°. AUP oe Ne SOV a ; oe es = / oe 7 \PERU 1) Ko PoRTION OF PRov. SANTIAGO—-ZAMORA y REPUBLICA pvEL ECUADOR Nj y Sources of data:— ¥ % Primary reference points Amer. Geog. Soc. |: mill. map. Rios Namangoza, Zamora, and Yunganza — Sketch and notes, Mistoneros Salesianos. Rios El Cruzado, Tintas and Yapi,and Indanza to Mendez — Field sketches and notes,H.Jorgensen. Rios Negro, Upano and Chiviaza — W.H.Camp. Collated by Camp,Jorgensené Prieto, Cuenca,/94$ ; Jel, WHC.,1957. OV PLA 8 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 and the valley bottoms may be mud flats, sand banks, or narrow boulder-strewn corridors, interspersed with massive, overhanging rocky ledges. The thing immediately obvious in this ever-changing series of habitats was the great abundance of species and the relatively few individuals of each. This is not true of the steep Andean escarpments, or for that matter of the outer slopes of the Cutuci., Duplicate specimens sometimes were a problem. Among the Rubi- aceae, the gaudily flowered Palicoureas offered no problem, since they are shrubs and trees. But the ubiquitous and somewhat more herbaceous Psychotrias some- times were most vexatious, for it seemed that no two were exactly alike. In the herbaceous members of other families one might see a small but seemingly vig- orous colony of some sort growing in the moss atop a boulder. There were hun- dreds of moss-covered boulders in the same area, all seemingly alike—but it was unusual to find another colony of the same plant. It was an amazingly complex flora and it hurt mightily not to be able to make a clean sweep of samples from it. But our job was hunting Cinchona. As I had anticipated, we did find Cinchona on the upper slopes and top of the Cutuct. Some of it was worthless; other samples gave quite good analyses, the latter in my opinion being related to the Calisaya type from much farther south in the Andes, But unlike the Cinchonas of the Andean slopes, which grow in ‘‘man- chas’’ or colonies, those of the Cutucdé occurred as occasional, scattered trees. In this they were consistent with the rest of the flora there. At the time I was deeply disappointed. On later reflection I was rather glad of it. The Jivaros would not have demeaned themselves by stooping to the type of labor necessary to har- vest the bark, and the importation of Cholo cascarilleros from the Andean uplands would only have caused bloody trouble here in the center of the Jivaro territory. One cannot leave this part of the narrative without at least further brief men- tion of the Jivaros themselves. At times they have been grossly maligned by those writers who have been on the fringes of their territory, and whose know- ledge of them is based largely on hearsay, by brief contact as they happen to come into the few places available to them for trade, or, worse yet, by those out- casts and loafers that congregate about the few missions. It is only after one actually has known the Jivaros of the eastern lowlands, hunted with them, and lived for some time in their homes, that their true character is evident. They are a proud and stiff-backed race and, so far as my knowledge goes, have never been conquered or subdued. (The much-publicized ‘*Colorados’’ of the lowlands west of Quito are a different race, now drunken and debauched.) The Spaniards early founded cities in the Oriente, expecting to make slaves of the Jivaros, either by force ot by debauching and degrading them. Neither of these methods succeeded and, ultimately, the Spaniards were forced to withdraw. Tradition has it that the present village of Sevilla de Oro (which will find a place in a later section of this narrative) was named as a memorial to the Sevilla de Oro which the Span- iards founded in the Oriente, but which was wiped out by the Jivaros. The Jivaros are excellent gardeners. In fact, I have rarely seen cleaner or better stocked gardens among the so-called primitive peoples of the Americas. The basic food items are the sweet potato, yuca, and maize, furnishing both pro- teins and carbohydrates. Several types of sweet potatoes are raised. The yuca is the ‘*sweet’’ type, that is, the selected form devoid of the bitter, poisonous prin- ciple in the form grown widely by the ‘‘down river’ tribes, and which must be grated and washed in water before it can be used. A similar form of this plant is the source of our tapioca. Two quite distinct types of maize also are raised, one being primarily a parching type, the other sometimes ground into meal. The yuca, however, is the staple carbohydrate. Although the Jivaros are excellent potters 1952] PLANT HUNTING IN ECUADOR 9 and always have an ample supply of containers of all sorts on hand, they will not tolerate the soggy, pasty mess which the yuca root assumes when boiled, as it usually is throughout tropical Latin America; instead they steam it, whereupon it becomes a most excellent and tasty substitute for bread. The yuca root, however, serves another equally important service in the diet of the Jivaros. From it they make nijamang, their universal drink. The raw yuca root is first peeled, then chewed and spit into a large pottery vessel. It is inter- esting watching an expert at the job, for it is almost a continuous process. A large bite is taken from the root, chewed rapidly for a little while, the excess water from the root being rolled out of the mass with the tongue and expelled from one corner of the mouth, while the chewed mass, by now thoroughly mixed with saliva, is then ejected into the ready container. When about full, it is set aside. After some hours a little of the previous batch is mixed with the new and the con- tainer is again put aside for a few days to ferment. After proper fermentation, as much as is immediately needed is mixed with water and drunk. The biological principles in the preparation of this combined food and drink are simple. The starch of the yuca root must first be converted to sugar, and the only enzymes easily available to them are those contained in their own saliva. The bit of a former batch already has in it the yeasts necessary to begin the con- version to alcohol. And the growing yeasts supply the much-needed vitamins in a diet otherwise low in these necessary items. If I were to presume to offer a bit of advice, it would be this. The first time you visit a jivaria, you will be offered a large bowl of nijamang. Don’t sip it. Drink it without pausing for breath, then smack your lips volubly at the end and nod approvingly. Not to do so is to offend your host right in the beginning—and from thefe on you will be a social outcast and encounter indifference and obstruc- tion. It is also a good idea to distend your stomach at the same time and pat it satisfiedly, otherwise you will be handed a second bowl. From there on, you will be on your own for, if properly matured, it can be insidiously heady stuff. Also, until one becomes an old hand with it, a too free indulgence produces about the same effect as an over-dose of epsom salts. Anyone with a queasy stomach had better stay out of the Jivaro country. Their gardens also contain other food plants. A species of Marantaceae is quite often seen and one of the Cannas also is used. This latter serves a double function; its fleshy rhizomes may be eaten, but more often the leaves are used to cover food utensils or to wrap packets of food for a journey. The papaya also ts commonly raised. I am fond of this tropical fruit and on various occasions tried to buy one. They would give me all the green ones I wanted, but they would neither give nor sell me a ripe one. They regularly stew the green papaya, but simply will not use them when ripe. Thinking that perhaps, for some reason, the variety raised by them was inedible when ripe, I surreptitiously took a ripe one and sampled it. It was by no means as delicious as those purposely selected for rich- ness of flavor when ripe, but it was quite acceptable. I never got to the bottom of this taboo on ripe papayas, but there must be some reason for it, otherwise it would never have come into being. The plantain, or cooking banana, has been introduced, but has not made much headway. I join with the Jivaros in deciding that, properly prepared, their yuca is vastly superior in flavor to the plantain. Also yuca stores much better than the plantain, has much less waste, and is much easier to transport on long journeys; furthermore, yuca can be made into nijamang, whereas the pulp of the plantain would rot and become a putrescent mess before the fermentation process could be accomplished. 10 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 Where barbasco is not naturally abundant, the Jivaros also propagate and raise this fabaceous plant. It is used as a fish poison. Pieces of the plant are beaten over the edges of their dug-out canoes (for usual river transport, they prefer the buoyant and easily constructed rafts). When sufficient material has been pre- pared, it is then mixed with water in the bottom of the canoe (usually by treading) and allowed to “‘steep’’ for a while. The canoe is then paddled to the upper part of a pool and tipped over. They then right the canoe and paddle to the outlet and select the fish they wish to keep as they come floating past. Obviously, fish of all sizes and ages are killed. They are well aware that a too-frequent poisoning of the fish seriously depletes the stock and so this is done only at relatively rare intervals. I was fortunate to see the process. Much of their actual fishing is done on the smaller streams where they have developed a most ingenious trap for use among the rapids. This trap operates by gravity. It actually strains the larger fish out of the stream, places them in the dry part of the trap, and so kills them with- out the need of anyone being in attendance; the small ones and fry are automati- cally dropped back into the stream to develop further. Other plants are grown. One of these, a vine I suspect of being Apocynaceous although I did not find it in flower or fruit, is trained on special arbors only in the gardens of the head-men; it is used in some manner in certain of their cere- monies. Tobacco also was seen, but I could not make a guess as to its species since, at that season, the plants were not yet in flower. In certain gardens one also finds beautiful specimens of an arborescent Datura; these will be briefly dealt with in a later passage. A regular sight was a small plantation of the native perennial, arborescent cotton, raised by them since time immemorial for its long fibers. The Jivaros are excellent weavers, and have evolved a curious but effec- tive loom. The cloth they produce is sturdy and fashioned with subdued but pleas- ing color patterns made from native dyes. And every garden has a stump covered with a species of Manettia; the specimen I brought back was presented to me by one of Patéhi’s wives in return for having treated a jungle sore on one of her children. The plant bore masses of delicately tinted pink flowers. It is their rem- edy for dental caries; rather, I should say their preventive. Many dentists think that it is a ‘‘sweetened’’ tooth rather than a dirty one which is prone to decay. The constant chewing of the yuca root leaves particles of its starch between the teeth, and these would be fermented to sugar by the salivary enzymes and so lead to dental caries. When I returned to the States, leaves of this specimen of Manettia were tested in the laboratories of the New York Botanical Garden for their ‘‘antibiotic’’ or bacteriostatic activity. The dried material was not found to be particularly active on the organisms tested. Dentists are ndt completely agreed on the exact cause of tooth decay and it may be that the leaves of this plant are effective in some manner not understood. All I know is that after chewing a batch of nijamang the Jivaro women regularly go to the garden, pluck a few leaves of this Manettia, and chew them. They also stoutly maintain that it is the only thing that keeps their teeth from decaying. The Jivaro matrons may become progressively more black-toothed as they grow old, but they do not become snaggle-toothed hags in their late twenties, as do the women of the upland, Andean tribes. _ Jivaro marriage customs are interesting. The tarimiat, or first wife, usually is bargained for between the parents when the future husband is in his ’teens. There- upon he goes to live with his future in-laws. If, at the end of the first year of the betrothal, the girl is not pregnant, the wedding is called off and he goes back home. If, however, she is with child, the actual ‘‘wedding’’ ceremony is still some time off, for he has not yet proved himself capable of taking his full place in the 1952) PLANT HUNTING IN ECUADOR ib community. First there is a j@a or house to be built. This is no simple, three- sided jungle shelter. It is a large place, with a series of separate interior apart- ments, for himself, his first wife, the anticipated wives to follow, and their broods of children. Arrangements also must be made for the cooking place for the whole household, a general space and benches for conversation when visitors come, and places where they may sleep when staying overnight. This all is under one roof. The timbering must be well engineered for so large a structure, and the palm leaves for the great expanse of roof thatch individually and carefully smoked over ‘a special fire so that they will resist rot. The prospective bridegroom now has his tarimiat and his jea but, according to tribal custom, he still cannot rightly claim his place in the social structure, and he may not occupy his house. He first must prove his prowess in combat and ap- pear with a tzantza of his own. Since childhood he has been practicing the art of making shrunken heads—and it is an art—by using those of monkeys caught in the chase, and so this is no new thing. But this may be his first human one. Also, one doesn’t just wait along the trail and plunge a spear through the first unwary passerby. It has to be obtained some distance away and in enemy territory, either as part of a regular war expedition or as a lone-wolf affair. Having taken his head and made his tzantza, he may then return, enter his house by the front door with full ceremonies, and take his place in the community as the head of a household or jivaria. From there on, additional wives may be acquired either by barter or theft. The details of the making of the tzantza might be of some interest. Also, one could record the complex social customs of these interesting people. But these are outside the bounds of a brief note on the collecting of plants in this remote region. One might add, however, that some of those who have writter what seem to be authoritative works on these subjects clearly indicate that they have never lived with the Jivaros even for a brief time and so have no real understanding of the situation. For example, three wives are a minimum number for a well-ordered household. There must be expeditions away from home for hunting and fishing to augment the proteins in the diet. One wife must stay at home to tend the garden and take care of the accumulation of youngsters. The other two go along to assist with the work. Usually they are nursing and must carry their infants with them. Whether the quarry be fish or monkeys, it is cut up and dried over a fire and smoked to further preserve the meat. Therefore one wife must remain in camp dur- ing the day to tend to these chores; the other wife goes along to assist with the fish traps or acts as a second pair of eyes, a great help—almost a necessity—in jungle hunting. The wife in camp also tends and nurses the child of the one out on the chase. The next day the wives exchange their duties. It is a system of complete cooperation in the round of family duties but, for its proper function, re- quires an imbalance of the human sex ratio, a 3:1 ratio rather than the usual 1:1 ratio. Those professional ‘‘do-gooders’’ who beat their breasts and deplore the taking of human heads by these people, do not understand the nature of the life required of them by their environment or the structure of the family. The elimina- tion of two out of every three males in the adult population is the only way in which the social structure can be maintained or the family kept as a functioning unit. The making of the tzantza is only a bit of symbolic ritual, closely akin to the ceremony of Communion as practiced by Christians—the taking of human flesh and blood as symbolized by the sanctified bread and wine, a ritual derived from the ancient and sacrificially bloody helioatric religions of the Mediterranean region. 12 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN (Vol. 8, No. 1 ° THE WESTERN ESCARPMENT Early in April, 1945, I was called to the Quito office to close up the work of the explorers of the Misién de Cinchona. The others already had left, some of them several months previously. There was a last hurried errand along the Rio Pastaza into the Oriente east of Bafos. And then on April 18 the work was of- ficially ended. The next day I set out on the return trip to Cuenca, which was to be headquarters for the remainder of my sojourn in Ecuador. My old crew of Jorgensen, Prieto and Giler reassembled the last days of April. I had rented the residence which had housed the bodegas of the old Misidn de Cinchona in Cuenca. This was admirably suited to our needs, for it gave us ample room, both for the storage of specimens as they accumulated and as a residence for the crew when in town. At last the collecting could follow a predetermined pat- tern and not be a casual adjunct to the necessarily primary activity of search for Cinchona, As a result it was determined that our first major objective would be to attempt a transect of the western escarpment. Arrangements had been made previously with Dr. Herbert Spencer Dickey to occupy part of his establishment in the village of Huigra in the Cafion of the Rio Chanchan. We arrived at Huigra the evening of May 5 and the next day set up our equipment in the then unused hospital which Dr. Dickey once operated. Adven- turer, explorer, bon vivant, raconteur extraordinary, and tropical doctor, Dr. Dickey, until his retirement, had been chief medical officer of the Quito-Guayaquil railroad. His recent passing leaves a void among those choice personalities which one sometimes encounters in out of the way places. Active collecting in this general area extended from May 7 to June 19. During this time, our activities took us out onto the coastal plain in the region of Naran- jito (Prov. Guayas) at an elevation of about 120 ft. Here Prieto became quite ill, for this was the first time in his life he had experienced real tropical heat al- though he had never been more than a few degrees in latitude from the equator. He was put on the train and sent back up to Huigra where I was certain he would recover. After a few days in the region around Naranjito, we moved back toward the mountains, to the town of Bucay. This area is at about 1000 ft. elevation and still quite uncomfortable to one who has spent the previous year at much higher elevations among the Andean peaks. Jorgensen fell ill soon after we arrived in Bucay and was sent home to Guayaquil, where he could get adequate medical at- tention. At this juncture it was ascertained by telegram that Prieto was recover- ing rapidly and able to operate the driers. Therefore we sent the bulk of the equip- ment toghim, at the same time making arrangements for the bundles of undried collections to be put on the train early each morning and unloaded in Huigra, where Prieto would be waiting for them. This left Giler and me completely free to devote all our energies to collecting. They were small enough, since both of us were having attacks of fever, which did not make the work easier. We stuck it out for five days of collecting in the region around Bucay and then, finding that the law of diminishing returns had begun to catch up with us, returned to Huigra. Prieto had been almost swamped with specimens, and so, while we all got needed rests, we cleaned up the last of the accumulation and those ‘‘stubborn”’ semi-succulents which always give difficulty in drying. Jorgensen returned but, like the others, was not in good shape. There still was a stretch between the 2,000 and 3,000 ft. elevations which we had not touched. Of the three, Prieto was then in best shape and so on June 19 he and I set out before dawn, feeling our way across the ties between the rails. We arrived at the 3,000 ft. level just 1952)-» - PLANT HUNTING IN ECUADOR 13 as the sun came up, and from then on it was a mad scramble up the walls of the cafion and into the lateral valleys for certain choice things which we had not col- lected before. That evening, just as the brief twilight slipped into darkness we passed the 2,000 ft. elevation, and later stumbled into the tiny village of Naran- japata. There we bought full-fare railway tickets and, as is the custom and some- times a necessity in Ecuador, hopped the first freight, strapped our heavy packs of specimens and ourselves to the narrow runway on the top of a freight car, and enjoyed the smoke and cinders during the wild, careening ride back to Huigra. With this, we had begun to make real headway in our transect. The crew was willing but in no shape to plunge into another round of heavy work, so I decided to return to headquarters for further rest and recuperation, and also to sort and pack the specimens in napthalene. Some incidental collecting also was done in the vicinity of Cuenca during that period. We returned to our project of the western escarpment on July 4. Contact was made with the highest vegetation zone worked in the previous period, and from there we collected onto the paramo. Perhaps as a matter of some interest, the material of this botanical transect of the western Andean escarpment was collected between elevations of 120 ft. and 11,500 ft. It consisted of an excess of 1,000 numbers of which, where pos- sible, a full set of duplicates were taken—usually six or more depending on the nature of the material. Although by no means complete, these 1000-odd numbers should give a fair sampling of the flora of the region at that time of year. We were now deep in July and I had wanted to see what the ‘‘winter’’ flora might be like on the Paramo des Soldados, west of Cuenca. Bad weather had postponed the trip several times. Finally we set out but were turned back from our objective by a snowstorm. Rather than call the day a total loss, we tried a small valley near the edge of the paramo. July 16, spent in the region of the tiny glacial lake in the head of the valley of the Surucucho, is a memorable one for, as we left the valley that night, our pack animals were laden with more than 550 specimens. This single day of collecting has been chronicled elsewhere (Jour. N. Y. Bot. Gard. 47:25-31. 1946) and need not be detailed here. An opening with permanent employment and chances of advancement came to Jorgensen. He offered to stay with me for the remainder of my work, even on the chance of losing the position, but I felt he should accept immediately. Although we had a little farewell party it was not a very hilarious affair. THE, CORDILLERA ORIENTAL With time running out on me a decision was necessary. There were so many places which might be collected that it almost was a temptation to make a sort of grand tour, skimming a few trophies as we went along. There was even a tempta- tion to go into the region of Chimborazo and collect around this great snow en- cased volcanic cone.-It was decided otherwise. Almost every previous plant ex- plorer had ascended this mountain and so I determined to turn elsewhere. Being more accessible, both the western and central cordilleras had been earlier ex- plored by others. The eastern cordillera is botanically but little known except in a relatively few areas—and these from rather scant materials. Furthermore, in my work with the Mision de Cinchona, I had long noted the botanical richness of the sotobosque—that zone between the high forest of the humid regions and the tree- less paramo. On the eastern cordillera the sotobosque often assumes the char- acter of a true, high altitude “‘mossy forest.’’ As it breaks onto the paramo, the sotobosque often becomes dwarfed and there is replete with shrubby forms and herbaceous materials. Under the best of weather conditions this is a zone of 14 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol. 8, No. 1 almost constant fog and we were now getting into the worst of the rainy season for that part of Ecuador. Trails onto and crossing the paramos of the eastern core dillera are relatively scarce. We therefore decided to try and find a place some- where near the little village of Sevilla de Oro to set up temporary headquarters for our work in that region. A place was found in the mud-walled house of Sefiora Nieves-Cordova, es- poused to one Jesus Villavicencio. The house was located approximately 3 kilo- meters north of Sevilla de Oro at an elevation of about 8,000 ft. It was the highest habitation in this fog-drenched region and admirably suited to our purpose, for preliminary work on a projected road to Mendez had opened a means of access from Sevilla de Oro along the shoulder of the valley of the Rio Collay toward the Rio Paute, above the cultivated and pastured zones. From this, we could pene- trate the last of the high forest and soon reach the sotobosque zone; on occasion, we found obscure trails which led to the paramo, there somewhat above 11,000 ft. Both the old and new trails from Sevilla de Oro across the Paramo del Castillo also were available for trips to the higher elevations. Active collecting was begun in this area on July 27 and continued through September 4. As I have said, it was the height of the rainy season, and there were days on end when we never saw the sun; rarely did we return at night without having been drenched at least once. In general, however, the rains did not come early in the morning; usually we already were so far along the trail that there was little use in turning back, and so collecting continued for the day. I was frankly surprised at the number of plants in full flower during this seem- ingly inauspicious season. The bulk of them were entomophilous and I almost wondered whether the insects that pollinated them might be aquatic. It soon be- came evident that many of the Andean insects of this region were a special type that did not wait for sunny weather to go about their chores, but would emerge from their hiding places as soon as the rain slacked, and worked apparently with full vigor in heavy fog. On those brief periods when the sun did shine, the air was alive with the hum of the more timid sorts as they seemingly attempted to make up for lost time. Mention of flower pollinators leads inevitably to the hummingbirds. The Andean hummingbirds never ceased to amaze me. I had read of them, but one has to see them to appreciate their great variety. In size they run anywhere from something no bigger than the last joint of one’s thumb to as large as blackbirds; they have borrowed almost every color in the spectrum’s range and magnified them with brilliant hues, But it is not their size or color that astonishes, it is their probing bills. Some turn down and some turn up, and some point straight ahead, some of these ‘with bills so long that, in flight, they seem like jet-powered bodkins. Each has evolved a certain form of nectar-probing bill that enables it to work ona special type of flower, and to which it is limited. One day I saw one flying along which seemed incredible. Even for an Andean hummingbird it was fair-sized; shortly after leaving the head the bill turned upward at almost a 90° angle. It flew to a clump of Crimson Angel’s Trumpets scarcely more than ten feet from where I stood, hovered beneath an open flower, and then slowly raised itself on its rapidly beating wings until the bill reached the nectar deposits of the flower. When I say that the narrow corollas of. Datura sanguinea are about 10 inches long and that the nectar is located at the base of the flower (the flowers hang down), one may gain some idea of the astonishing, periscope-like nectar gathering apparatus of this bird. And mention of this plant also leads to a brief note on the group. Datura san- guinea is supposed to be a native of Peru. It is widely cultivated as a roadside 1952] PLANT HUNTING IN ECUADOR 15 *‘living fence’’ in parts of the northern provinces of Ecuador; I never found it thus in the southern provinces. I did find it in the southern provinces, but only about old temple platforms or ancient ruins. The Jivaros of the eastern lowlands regu- larly raise other kinds of this genus, these being perhaps somewhere in the gen- eral affinity of the D. arborea group, but usually in soft pinks and salmony pastel shades. As individual plants in the gardens of these head-hunters, they make a striking show, and far outrank in beauty any hybrid Daturas I have yet seen in cultivation. It was with deep regret that I could find no seed on certain of these plants (they may be self-sterile); and it was impractical to attempt to make cut- tings at that time. The seeds of the group are supposed to contain a potent nar- cotic. However, it is the leaves that were and still are used. Jorgensen told me how, while he was panning for gold in the Oriente some years previously, an accidental gunshot lodged in the muscles of his leg. No doctor was available, yet it was necessary to remove the bullet if he were to re- cover properly. The Jivaros took charge of him, made a decoction of the leaves of the species which they raise for this purpose, and had him drink it. He soon be- came drowsy. When he awoke (they told him that it was about 36 hours later) the bullet had been removed and the deep cut skillfully poulticed with native medica- ments. Unfortunately, he could not see. He complained of this, but was told to be patient. He said that he did not regain normal vision for another four or five days, after which there was no apparent further effect of the narcotic. One of the common sights in archaeological museums is the ancient skulls of former inhabitants of the Andes, often with holes where they had been quite skill- fully trepanned, probably for the relief of abscesses of the brain or similarcerebral afflictions. Many give evidence of post-operational healing, indicating that the patients survived and lived for quite some time afterwards. In speculating on these items some have wondered how they held the patient still for so tedious and painful an operation. I think we need hunt no farther than the genus Datura for the anaesthetic which these ancient surgeons used. The common occurrence of these trepanned skulls, especially in the older archaeological sites, has puz- zled many workers. They might be reminded that syphilis is endemic to the Andean highlands and that the inhabitants have been afflicted with it so long that they have evolved a race now apparently almost immune to its secondary and worst ef- fects. In my year and a half in the back country of Ecuador I never saw a case of what I would suspect was syphilitic paresis among the true natives. With the in- digenous inhabitants such an infection is scarcely more troublesome than is the common cold with us. Their favorite ‘‘cure’’ is a decoction of the native Ephedra. But I wander from the field of plant exploration into ethno- and medical botany; I trust, however, that someday my fairly extensive notes on the native pharmaceu- tical plants, as employed by the Quechua-speaking peoples of the region, will be collated with the plant identifications yet to come and so made available to work- ers in this specialized field. THE COLLECTING In the year and a half of field work in Ecuador, something over 5,200 separate numbers were collected; with the duplicates, materials for approximately 26,000 individual herbarium sheets were prepared. To these also should be added certain small collections of my three major assistants, taken when they were on special assignments. In general, however, their specimens, although credited to them on the tickets, were run into my own number series and so appear in the general tally. As intimated in earlier passages, during the first year, while employed in the search for Cinchona, the collecting was desultory, or connected in some way 16 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol. 8, No. 1 with that exploratory work. Also, during that time when naught but oddments could be snatched along the trails, or where an occasional day only could be devoted to straight botanical work, the duplicates were not so ample as one might wish. The last six months saw a much better organized system of collecting and hand- ling the material; this resulted in our ability to produce much larger numbers of specimens. I have been asked on occasion for details of the methods we used and so, for the sake of those who may be projecting a fairly substantial collecting expedi- tion, these may be outlined. In the first place, it is useless to think of collecting either efficiently or on anything like a satisfactory scale in the rainy tropics un- less one dries the specimens with artificial heat. This is a prime requisite. I have discussed this item at greater length elsewhere and need not repeat the arguments in favor of the system (see: ‘‘On the Use of Artificial Heat in the Prep- aration of Herbarium Specimens,’’ Bull. Torrey Club 73:235-243. 1946). I would, in fact, extend the method there outlined to any part of the world, and to any type of plant material, if there is need to handle any but the smallest amounts of ma- terial, and if one has a sincere wish to turn out quality specimens. So far as the actual collecting is concerned, I would not be caught with a vas- culum as used in the usual manner. They are an utter abomination and have re- sulted in more poor specimens than one wishes to contemplate. If one must use a container, then use a waterproof bag of some sort. But in doing so, take special care that the material of each number is wrapped separately (usually in a piece of newspaper) and so placed that odd flowers and other parts will not fall out and become mixed with other specimens. ,The argument that one saves time by jamming a lot of loose specimens into a vasculum, to be sorted that night, is fal- lacious. Such statements are made by those who have never done any time-studies on the necessary operations to produce a dried and finished specimen. In actual practice, my usual method is to collect directly into the newspapers, using only an occasional blotter between every three or four numbers, if they have ample duplicates. The actual field procedure is as follows. The field presses are loaded with the amount of newspapers one soon learns will be necessary for the day’s work, plus a little extra for emergency collecting. These presses should have special straps so that they may be used as packs, or special slings can be fashioned according to the carrying customs of the assis- tants available. (The tump-line over the head is favored by most indigenous car- riers, but I personally prefer a combination of both tump-line and shoulder webs.) Arriving at the scene of the first collecting station, the presses are shed, along with other equipment not then immediately needed. Sometimes one assistant would be assigned to a particular task, the climbing or felling of a tree, or to the col- lecting of a particular series of plants. More often, as our group became experi- enced, we would split up and individually go in different directions, each agree- ing to stick rather well to a certain type of habitat. It also would be agreed that we would return to the presses at a certain time, usually at the end of a half hour or so. Each man carried a machete, and the first man back would enlarge the clearing a little if it was not sufficient in the first place. He then would lay his specimens out in small piles, carefully separating the individual collections. Soon we all would be back and the actual process of putting them away would commence. Each man would stand beside his piles of specimens and I would have my small field book ready with the first number. The number would be placed promi- nently on a sheet of newspaper and the assistant quizzed about the details of the plant—its type, habitat, or any other item not visible on the specimens. Then, 1952] PLANT HUNTING IN ECUADOR iz with the fresh material before me I would add other pertinent notes, such as color of the flowers, etc., etc. The newspaper with its number outside then was placed on the ground. For woody materials, I always carry a pruning shears in a special holster. This is then brought into play and as many specimens cut from the ma- terial as needed and placed on this individual sheet; extra flowers or fruit, if desirable, are then stripped from the remainder of the collection, and the residue tossed out of the way. We would then go on to the next collection and repeat the note-taking process and the trimming of sheet-size specimens. In the meantime another assistant would follow behind with a bundle of news- papers and start at the first pile of specimens, already trimmed and ready for the papers. These would then be placed individually in their papers, using enough to clean up the pile, any odd flowers or fruit being scattered through the collection of sometimes given a separate sheet of their own. The paper first placed on the ground, and which bore the number of the collection, was then folded around the whole set, so that the open margins of the other sheets would be closed. This packet would then be laid lightly in the press, or held down merely by the weight of a small stone or some other object, in case there was any breeze stirring. Ar- riving at the end of the first assistant’s collections, the one then putting the ma- terial in press would come over and we would start out on his collections, the first man working on the specimens where the other had left off. My own collec- tions would be similarly cared for. In this way, in a remarkably short time, we had converted the lots from raw material into specimens, already cut to herbarium sheet size, and filed them in their newspapers; the field notes also had been com- pleted. The individual packets of collections were then placed in press in serial order, as they had been numbered, with an occasional blotter interspersed when the packets became a little too thick; the blotters are not necessary, but serve to keep the contents of the field presses on an ‘‘even keel’’ when opened at the next stop. Soon we would be on our way and ready for the next collecting station, where the process would be repeated. In this manner there was no packing of a lot of excess waste material back to camp, ‘‘to be cleaned up when one has more time that night.”’ In a well run plant explorer’s base camp, there isn’t any extra time at night for this kind of messy fumbling. Also, it is worse than foolish to trust to one’s memory about the details of the plants of the day and jot the notes down that night. There have been more errors perpetrated by this sloppy method of assembling field notes than many are willing to admit. The field notes should be taken in the field, and not jotted down hours later (or sometimes hastily the next morning) after one’s memory has be- come hazy about certain details of height, habit, and habitat. Our drying stoves also were our cooking stoves. Therefore, immediately after supper, the residuum of the material on the fire the night before would be sorted and any completely finished pulled out of the presses. In the meantime, the ma- terial which had been in press for a day, but not on the fire, was being sorted and the plant materials carefully arranged on the sheets. (In the paper on artificial heat, I stressed the desirability of ‘‘seasoning’’ the specimens for a 24 hour period in blotters before putting them on the heat; this accounts for the supply of plants to go on the fire being at hand soon after supper.) These ‘‘seasoned’’ specimens, now carefully arranged, would then be placed on the fire. Then the collections of the day would be tackled. | By that time, those which had been collected earlier in the day would have been somewhat ‘‘tamed’’ and ready for their preliminary arrangement on the sheets. However, there was an additional chore. Before any arrangement was done, each sheet would be numbered with the number on the covering slip-sheet. This is one 18 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 operation I discovered should better be done in camp at night. The reason is rather simple. We often collected in the rain. As a result, the sheets sometimes were soggy. Paper was too scarce in Ecuador to be thrown away and so these soggy sheets were laid aside to be used again and dry ones substituted. Had they all been given a number in the field there is a possibility that the number might not have been crossed off when the paper was used the second time, and so cause confusion of numbers at some later date. If the sheets were only damp as taken out of the field presses, they were used that way. After considerable experimenta- tion I have found that a special but not expensive item known as a ‘‘surgeon’s skin-marking pencil’? is the best. A soft and waxy pencil is desirable, but it should not be friable or gummy; the skin-marking pencil has the proper texture and also will work on fairly wet paper, making a heavy mark without tearing. In putting the plants of the day’s collecting into press, it is not necessary to have completely dry blotters. In fact, it is my opinion that with many materials, slightly damp blotters are best for the first 24 hours. The main purpose of this **seasoning’’ period, as I see it, is to permit the plants to carry on a little anaer- obic respiration—as they will in a tight and slightly dampish press. This results in the conversion of a portion of the carbohydrates to fatty compounds and these, in turn, seem to distill throughout the plants while on the heat at a later period and so keep the specimens a bit more pliable. As they become too damp, the blotters may be fed into the driers with the specimens. By using these systematic methods of collecting and attending to the speci- mens, three of us—Prieto, Giler and I—tumed out some 6,600 specimen sheets in about 40 days while in the eastern Cordillera. Part of the time Prieto was ona wild-goose chase across the range in the Oriente trying to track down the flowers of several species of Cinchona I wanted but which we had missed the previous year, Giler was sent back to Cuenca on various occasions with accumulated ma- terials and to replenish our supplies, population-sampling at different altitudes in several groups occupied various days, and we soon cleaned up the area near headquarters and so often had to climb miles at altitudes near or above 10,000 ft. to get to really productive new areas. And all of this was done during the height of the rainy season under the worst possible climatic conditions for field collect- ing, and at a season when the flowering material was at its low ebb. I have re- counted the foregoing not in a spirit of boasting, but to indicate the reasons why I am a firm believer in systematizing the collecting routines and also why I ama strong advocate of drying by artificial heat and the use of metal corrugates. With all of this we also found ample time to visit back and forth with our new Cholo friends, and make merry when the occasion arose. It was not a dull time with all work and no play; it was a full and satisfying experience. In the meantime, al- though we did not then know of it, a bomb had dropped on Hiroshima. ACKNOWLEDGEMENTS One cannot properly acknowledge the help rendered in so many ways by so many people in work spreading over so long a time and under such varying con- ditions. The officials of the Mision de Cinchona in Quito were most encouraging. I have already mentioned the work of my three main assistants. One also remem- bers those times when mule’s would flounder in the mud of the trails, and when the muledrivers would have to jump into mud to their belts and hold up the heads of the mules to keep them from smothering, while others jumped in and quickly cut the pack ropes to rescue cargo boxes of precious specimens, even before they extricated the mules; the mud does get deep on Ecuadorean trails. And the FO5a} > PLANT HUNTING IN ECUADOR 19 thousand little courtesies extended to a stranger as one would go about one’s daily tasks. None of these can be properly acknowledged here. I would also wish to thank the Director of the New York Botanical Garden for having obtained a grant of $2,500 for the work of the last six months. It was an opportunity which I felt should not be wasted, since I knew something of the coun- try and had available a crew of excellent assistants already trained in the details of plant collecting. But to maintain this crew, even at very modest salaries, and to take care of our daily living expenses and travel, it was necessary to add to this sum an equal amount from my own funds. I have never regretted this personal investment toward a better understanding of the complex flora of this part of the American tropics and also to fill out my botanical education. Also, I would wish to thank my former colleagues of the New York Botanical Garden who undertook the care of the collections when I left that institution, and especially Mr. John Wurdack, onto whose shoulders fell the usually thankless and always tedious task of sorting the bulk of the collections and of seeing to the typing of the labels, chores which were only partly done when I left. Lastly, it is my great pleasure to thank those various specialists in plant taxonomy who have studied and still are working over various groups of this ma- terial. I naturally await their determinations with considerable interest. Some of the specimens doubtless will be from widespread Andean ‘‘weeds’’; I often knew them to be so, but even ‘‘weeds’’ are often of ethno-botanical importance. Others of the specimens are certain to be from little-known kinds, and so increase our knowledge of them. And a few, perhaps, may prove to be those cherished jewels of all taxonomists—the so-called ‘‘new species.’’ And here it should be admitted -that, as a person primarily interested in the genetic structure of plant popu lations, I was not above slipping into the collections of groups other than those of partic- ular interest to me certain series of “‘intergrades’’ between what I supposed might otherwise appear to be rather sharply defined species. This was not done to fur- ther perplex my taxonomic confreres, but as a possible aid in furthering an under- standing of the complexity of our tropical floras. I trust that, in general, my field notes will serve to indicate the nature of such collections. Several technical papers in part based on materials collected during the course of this work already have appeared. They are as follows: Camp, W. H., Cinchona at High Elevations in Ecuador. Brittonia 6:394-430. 1949. Steere, William Campbell, A Report of Some Recent Collections of Rubiaceae from Ecuador. Bull. Torrey Club 72:295-311. 1945. COLLECTION NUMBERS On leaving for Ecuador, there was no idea that any sort of extensive collect- ing was to be carried out; certainly the work of the last six months was not en- visioned. The records of my previous collections had been stored in New York and the boxes moved out of my former office. It therefore was impossible to as- certain the last number in my regular collection series and I had forgotten it in the many activities of the early war years. Therefore it was necessary to Start a new series; however to avoid duplications, these were prefixed with an ‘‘E’’ (indicating Ecuador) to distinguish them from the same number of my earlier collections. Certain of the collections made during the work of the Misién de Cinchona _ have somewhat complicated code letters. This was advisable because of the need 20 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol 8, No. 1 for correlating the analyses of Cinchona barks with a series of field collections of herbarium materials through the medium of chemist’s reports. To avoid cleri- cal errors in the Quito office, a special system of numbering was developed for the various small exploring parties under my direction. It was almost certain that errors would creep into so large a series and also that seeming inconsistencies would develop. A few may be noted which already have come to light and are here appended. 643. **Valley near Portovelo.’’ This is the valley of the Rio Amarillo. 700-752; 1620-1644. ‘‘Chaparral and paramo E. of El] Pan.’’ This is the Paramo del Castillo of later collections. : 1392. Apparently there are two collections with this number. 1682-1710. This material, collected in May, 1944, in the Oriente in the region of the Rio Pastaza while in company with Dr. W. C. Steere, was not then given serial numbers. It is therefore chronologically out of place in the series. 2296-2348. In an unfortunate transcription, these numbers were dated April 26; they were collected March 26. NOTES ON OBSCURE LOCALITIES In general it was my aim to locate the collections while in the field so that anyone wishing to find them on a map (or perhaps revisit the area) might do so with ease. To this end the American Geographical Society made me a special fold- ing field map, assembled from various sheets of their usually excellent map of Hispanic America, 1: 1,000,000 scale, the Provisional Edition, then current. This was my constant field companion for a year and a half and is beside me as I write these notes. This provisional edition doubtless soon will be revised. However, its scale was such that, on occasion, it was necessary to use local maps with more detail. Lacking these, as in our work in the Oriente, it was requested by the office of the Misién de Cinchona that we prepare such reconnaissance maps as would be needed for possible future work. These were necessarily sketchy, since we had no proper instruments for traverse work. However, the map of this report, split through the center, has been taken from a larger map prepared and delivered to the Quito office on January 1, 1945. It is, so far as I am aware, the most de- tailed chart of this part of the Oriente yet published. The major parts of this map were put together from the notes and field sketches compiled by Mr. Jorgensen and myself with, of course, such local information as we deemed reliable. Ade- quate indications of latitude and longitude have been omitted; key points will serve t6 orient any user of these charts when correlated with a standard map. The permanent labels for the specimens were printed on my return from Ecua- dor. Where the numbers of specimens warranted, the localities were printed di- rectly on the labels, leaving only the field data of each collection to be added. In doing this every attempt was made to include such information as would serve to locate some of the more obscure places not appearing on the American Geographi- cal Society’s maps, by reference to some well known and easily located place. This was not always possible. Therefore for the sake of those desiring as nearly exact localities as possible for mapping purposes, a list of those which I deemed might give most trouble has been compiled. Where the Society’s map is referred to in the list, it is abbreviated as ‘tthe 1/m map.’’ In the Oriente, transliteration of Jivaro place names gave considerable trouble. I am not a student of linguistics, and it was only after the printing of the labels that I had opportunity to examine in any detail the work of P. Juan Ghinassi— 7052] - PLANT HUNTING IN ECUADOR a1 *“Gramatica teérico-practica y vocabulario de la lengua Jibara’’ (Quito, 1938). This will explain the need for extended notes on certain place-names locating our collections in the Oriente. Jivaro is not a simple, ‘‘primitive’’ language; it has a highly complex grammar and is replete with fine nuances of meaning depending on inflection and the differences in word endings. The altitudes and distances are sometimes given in metric and at other times in English units. Where Ecuadorean maps or the Society’s 1/m map were used as a basis of reference, the distances and altitudes are in the metric system. Where distances were taken along roads during travel by Jeep, the speedometer readings in miles were used. Of the various altimeters available to me, not one was cali- brated in meters; therefore to avoid errors incident to conversion, the altitudes in feet were recorded in the field and carried through the notes and onto the labels. For some reason, passing storms in the equatorial Andes do not produce the marked effect on barometric pressures which one expects; this was noted as early as the time of La Condamine, in Andean Ecuador ca. 1735-1743. Therefore, if properly calibrated at some known point of reference from time to time, field aner- oids are amply accurate for the purposes of plant collecting in this region. The alphabetized list of obscure place names, or those which need amplifica- tion or correction follows. Alpachaca, Cordillera de—The Allpacha Silvan of the 1/mmap. The Pan American Highway (q.v.) heads roughly S-SW from Cumbe, follows the length of the Cord. de Alpachaca, and crosses the Rio Leon just north of Ona, thereby missing the towns of Nabon and Cochapata. Ambocas, Rio—Apparently misspelled as ‘‘Rio Ambarcas’”’ on 1/m map. ’ Bahos—Two places of this name appear on the labels, one in Prov. Tungurahua along the Rio Pastaza, the other in Prov. Azuay SW of Cuenca. Bucay—A town of considerable size on the Quito-Guayaquil railroad at the foot of the western escarpment of the Cordillera Occidental. It has replaced Huigra as the important place where the trains from Guayaquil are broken up and a few cars each are double-headed up the Cafion of the Rio Chanchan to Si- bambe and over the ‘‘Devil’s Nose’’ switchbacks to the uplands, where they are remade into regular size. Trails in the area follow no regular pattern; as a result we never knew just which province we were in since Bucay is lo- cated almost at the junction of Guayas, Bolivar, Chimborazo, and Cafar. In any event, for Nos. 3640-3851, all were taken within a day’s walk of Bucay. Cajanuma, Nudo de—‘‘Cross range’’ about 7 km. S of Loja, easily located on 1/m map by ‘*Paso Cajanuma.”’ Canillones, Tambo—Shelter house on new trail between Loja and Zamora, on Oriente side of range. Castillo, Paramo del—The paramo on the crest of the Cord. Oriental on either side of the trail from El Pan to Mendez (Map 1). The Oriente side of the paramo is marked by the starkly uplifted, almost spire-like dark rock which, _castle-shaped, gives the paramo its name; this rocky peak also is known as Cerro Negro. Carboncillo, Paramo de—On the Loja-Azuay border between Saraguro and Ofa. The 1/m map has the province boundary running up the valley of the Rio de Ona; a semi-official Ecuadorean map has it to the SW, following the divide between this river and the Rio Paquishapa. Cerro Negro, Tambo—Just below paramo del Castillo, Oriente side of range on trail from Sevilla de Oro to Mendez; Map 1. 22, MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vols 8, No. 1 Chasqui, Paramo de—Near the village of Chasqui, Prov. Leén. Chiguango, Rio—The headwaters of the Rio Yaguachi; the Rfo Chiguango has its primary source in the Paso de Mataperros. Chontal, Hda. (Hacienda)—Valley of the Rio Negro; Map 1. Chontal, Tambo—Above confluence of Rio La Paz with Rio Negro (same locality as Hacienda Chontal). Map 1. Chupiantza—Jorgensen Nos. OHJ 4 & 5; refers to uplands south of Mendez (see next). Chupiantza, Rio—There is considerable confusion over the application and spell- ing of this name. The 1/m map has the Chupianza (a variant spelling) just south of and confluent with the Rio Paute, being the lower reaches of the Rio Negro; I have used this in Nos. 1540-1570 as a locality and in conjunc- tion with the Rio Negro, so there should be no confusion. The Rio Chupi- antza of most labels is confluent with the Upano north of Mendez. To add to the confusion, the name Chupiangas, apparently interchangeable with Chupi- antza and Chupianza, is likely to occur on future maps, but whether for the stream north of Mendez or as the lower reaches of the Rio Negro will depend on the whim of the cartographer involved. See Mop 2 (also see above). Chupianza—See above. Chupiasa, Rio—Perhaps an error in my transliteration, since Ghinassi lists a Rio Chiviaza for the region. One night in camp the Jivaros (at my instigation) had an argument about place and river names for the region. As nearly as I] could gather, they agreed (with some dissenting voices) that this river was the Chupiasa rather than the Chiviaza; and so it was listed in my field book and on the labels. However, on the map I have bowed to Ghinassi, and used his transliteration as probably being more correct (Map 2). | Consuelo, Tambo—Between Rio La Paz and Rio Pailas; Map 1. Cruzado, Cerro el—Trail between Indanza and Gualaco; Map 1. Cumbe, Rio—Rises on paramo beyond Cumbe and is confluent .with the Rio Tarqui near the village of Tarqui. This area is enshrined in the history of Ecuador as being the deciding field of battle in the country’s fight for independence. Cutuct, Cordillera—Map 2. A range just east of the Rio Upano and about 80 km. east of the crest of the Cordillera Oriental, and a northward extension of the Cord. del Céndor. Its uplift was more recent than that of the Cord. Oriental, the most recently elevated of the main Andean ranges. Etzentza—See Itzintza. FF, CC.—Spanish abbreviation for Railroad. Gloria, La—Valley slope S of Rio Pastaza opposite El Topo (q.v.). At time of coliecdons this was a tree nursery operated by the Mision de Cinchona. Guagrauma, Nudo de—Cross range about 12 km. S of Saraguro and 55 km. N of Loja. Paso Ramos-urcu and Paso Atacana of 1/m map are in this range. Guyaba, Rio—Branch of the Rio Catamayo, entering it near La Toma; spelled Guyabal on 1/m map. Huatracaja—Uplands NE of Azogues; Map 1. This is the region of the Cerro YausAn (elev. 3632 m.) of the 1/m map. On several occasions while alone attempts were made to get to the top without cutting trail but I was stopped each time by a particularly heavy sotobosque developed near the 11,000 ft elev, Indanza—On trail S-SW from Mendez; see Map 1 and 1/m map. Itzintza—This should have been transliterated as Etzentza (Map 2). I listened with special care to the pronunciation of the name, for it was the locale of our base-camp in the central Cutucti, and to me it definitely nad more of an *?? than ‘‘e’? sound; however, I bow to Ghinassi’s transliteration. Patéhi 1952) PLANTS COLLECTED IN ECUADOR 23 explained that it was named so because it was the river where they got a certain kind of ornamental bead which they used (from Etza—the name of the bead, being the seed of a vine which I did not see in flower or fruit but prob- ably one of the Fabaceae, and Enza—meaning ‘‘water’’ or, by extension, **stream’’ or ‘‘river’’). Kumza, Rio—Confluent with the Rio Yunganza; Map 1. Leon, Rio—Rio Leonhuaico of 1/m map, NW of Ofa. Loma de Oro—Hill above 10,000 ft. elev. in the Nudo de Guagrauma (q.v.). Malacatos—About 27 km. S of Loja; on some maps listed as Valledolid. Not to be confused with the Valledolid of 1/m map further south across the Nudo de Sabanilla in the Oriente. Mataperro—The region around the Paso de Mataperros. Mataperros, Paso de—Cord. del Cisne, W-NW of Loja. Mirador—A locality on Cerro Partidero (q.v.) on main trail to Mendez. Moro-moro—About 10 km. west of Pihas, which is west of Zaruma on 1/m map. Naranjapata—A small village in the canon of the Rio Chanchan about midway be- tween Huigra and Bucay (q.v.). Narino, Dept., Colombia—While working on the uplands in Prov. Carchi, Ecuador, we inadvertently crossed the unmarked boundary and found ourselves in Co- lombia, near the village of Chiles. A few numbers were therefore actually taken in Colombia. Negro, Rio—Map 1. In Jorgensen No. OHJ 6 this refers to the lower Rio Negro - (or Chupiantza) SW of Mendez. Nudo—Literally a ‘‘knot’’ and applied to E-W ‘‘cross ranges’’ connecting the main, generally N-S Andean ranges; alphabetized under place name rather than ‘‘Nudo.”’ Ontza, Rio—Central Cutuct. Should read Rio Tzantza (Map 2). This is an error which occurs on labels for Nos. 1177-1197. The Jivaro with me at the time and who was naming the plants (and streams) as we passed was unusually soft-voiced and used a broad ‘‘a,’’ stressing the first vowel; I did not catch the low, sibilant ‘‘tz.’’ The original field book error was inadvertently copied when the labels were printed. Pailas—See Pilas. Pan American Highway—As used, this refers to the segment between Cuenca and Loja. The kilometers are numbered from Cuenca. Km. 20 is where the Cuenca- Giron road branches from Highway; Cumbe, km. 28; Paramo de Tinajillas, km. 45; Rio Leén, km. 97; Ona, km. 108; Paramo de Carboncilla, km. 125; Saraguro, km. 147; Nudo de Guagrauma and Loma de Oro, km. 159. From here to Loja the kilometers had not been marked. Pdaramo—Treeless areas at the higher elevations in the Andes, in Ecuador usually being above 10,500 ft. Alphabetized by place name and not under paramo. Parroquia—‘‘Parish’’; used by F. Prieto. These minor places have been located on labels with reference to places on the 1/m map. Partidero, Cerro—Just west of Mendez and south of the Rio Paute; Map 2. Partidero, EI—A locality with scattered farm settlements on the Cerro Parti- dero (q.v.). Pati (Jivaria of )—Ghinassi transliterates this common personal name as Patéhi, and it is so used in pertinent passages in this account. I listened with spe- cial care to the pronunciation of this name as given by ‘“‘Patéhi’’ himself, and wrote the above; what I did not then know was that the last syllables are often modified or dropped in informal conversation in Jivaro,. The name Pa- téhi means ‘‘second born.”’ Patos, Rio—Confluent with the Rio Negro; Map 1. 24 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol. 8, No. 1 Pilas, Tambo—Misspelled on some labels; should be Pailas. Valley of the Rio Negro opposite the Rio Patos, near the Rio Pailas. Map 1. Portovelo—Prov. El Oro, just south of Zaruma; incorrectly listed on 1/m map as Portoviejo. Pupazche—An area with habitations on the erosion benches between the town of Paute and the Huatracaja upland (q.v.). Puyo—A small village in the Oriente beyond Mera. Quebrada Gualacéo—A steep walled valley in the Cerros Huashinan south of Bl Pan. Saraguro—The spelling varies, certain semi-official maps in Ecuador apparently prefer this over Zaraguro. Sevilla de Oro—SE of the town of Paute, on the easter shoulder of the valley of the Rio Collay opposite El Pan (Pan of the 1/m map); Map 1. Sta. Elena—West of Mendez; Map 1. Sotobosque—Tangled and at times almost impenetrable dwarf forest, usually a typical “‘mossy forest,’’ at high elevations between the upper forest and the paramo, Surucucho—A glacial lake and valley W of Cuenca and surrounded by the Paramo des Soldados. (See Jour. N.Y. Bot. Gard. 47:25=31, 1946, for account of area and photographs.) Suscal—Directly west of Tambo. Tambo—Map 1. Until late 1945 the terminus of the branch of the railroad running south from Sibambe; track was then being laid on to Biblian. Also a rest house along a trail; where appearing as part of a name, it is not used in this index for alphabetizing purposes. For the most part tambos are ephemeral places, since the muledrivers have the habit of tearing them up for firewood. Many place names on parts of the 1/m map once were tambos, but no longer exist, Tayuza, Rio—Map 2 and 1/m map; confluent with the Rio.Upano just ‘north of Mendez. Tigre, El and Rio—Area of settlement where the Rio Tigre enters the Pastaza just east of El Topo (q.v.). Tinajillas, Paramo de—Upper part of Cord. de Alpachaca (q.v.) about 20 km SW of Cumbe. Tintas, Rio—A branch of the Rio E] Cruzado, in the Oriente SE of (El) Pan; Map 1. Future workers should know that this is the same area as that collected in by Dr. Julian A. Steyermark in 1943, probably under the locality name of the ‘‘Arenillas Region.” Toma, LamA small village west of Loja in the valley of the Rio Catamayo; actu- ally applied to the sandy, arid areas between the Rios La Toma and Guyaba above their confluences with the Rio Catamayo. Topo, El—A place along the Rio Pastaza east of Bafos, notable only as a station where trucks taking Naranjilla (Solanum quitoense) to the uplands pick up their major cargoes. , Tres Ranchos—West of Mendez; Map 1. Vilcabamba—About 38 km. S of Loja. This is listed on some ieee maps as. La Victoria. Yapi, Rio—Map 2, east of era: Cutuct, Yucal—A region N of the Rio Paute about 5 km. west of Mendez. Zapote area—On trail, Gualaceo to Indanza; Map 1. Zaraguro—See Saraguro. ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA Vol. 8, No. 1 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN March 26, 1952 PLANTS COLLECTED IN ECUADOR BY W. H. CAMP. BROMELIACEAE, CANNACEAE, ETC. LYMAN B. SMITH BROMELIACEAE Subfamily 1. Pitcairnioideae Puya aequatorialis André. Azuay: coarse much-branched plants, firmly attached to soil, leaves pale green above, silvery-scurfy below, flowering spikes to 2 m., the lower half bare except for scattered bracts; the first flowers loose on the spike, the upper crowded, sepals green under the scurf, petals green at base, a dark greenish- blue above, becoming reddish after anthesis, anthers bright yellow, ‘‘Achupilla,’’ very common on steep slopes; between Rios Azogues and Gualaceo, valley of the Rio Paute, between Paute and Cuenca, 7,200-8,000 ft., (dry cliffs, rocky hill- sides and occasional ravines), E-2322. Puya glomerifera Mez & Sodiro. Leon: in clumps, plants branched, in open, subprostrate, in chaparral, branches arched upward, flowering spikes to 1.5 m., flowers bluish-green; Paramo de Chas- qui, 12,000 ft., E-2349, Puya gummifera Mez & Sodiro. Azuay: single plants, basal leaves in dense rosette, flowering spike to 2.5 m., lower bracts subtending flower clusters leaflike to 20 cm. long, upper bracts 2=3 cm. long, thin, calyces basally green, apically yellowish, covered with brownish scurf, corolla a pale greenish ‘“‘isabellina”’ (i. e. a dirty white with greenish tinge), base of plant, especially the expanded leaf base, eaten by the common people and said to be ‘‘good for the kidneys,’’ also fed to cattle, pigs, etc., this probably explains why the plant is rather rare; along the Rio Cumbe, 25-30 km. south of Cuenca, 9,300-10,000 ft., E-2202. Leaves pale green, nitid, especially above, spines basally pale, apically brown, flowering spikes to 3 m., flowers creamy white with greenish tinge, favorite food for bear, stems also fed to pigs and ‘‘cuys’’ (guinea pigs); quebrada leading into the Rio Collay, 3-8 km. north of Sevilla de Oro, 7,000-8,300 ft., E-5198. Puya hamata L. B. Smith Azuay: inflorescence spikes bracteate and lanate, to 5 m., fruits on upper 2 m., leaves in dense rosette, ‘‘Achupalla-cimarona’’; although a conspicuous ele- ment of the open paramos, this species probably originally inhabited the occa- sional rocky declivities and cliffs of the region, where it is still present; cer- tainly it is never seen in undisturbed chaparral or slope-forest. With clearing, it moves onto the pastures and meadows, where it becomes a pest. This species dies after flowering and fruiting. The older plants seem to be resistant to the usual paramo grass-fires, although the younger ones perish. The usual method of control is to cut off the inflorescence at the time of flowering and before the seed is mature. When passing this area some months ago, the flowers were noted to be “blue.’’ As the fruit matures, the outer leaves become sharply reflexed, bending to the ground. So far as known, this area is the southernmost station for the spe- cies in Ecuador; at least it was not seen on any of the paramos south of here. 20 26 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vdl. 8, No. 1 fi i ZZ eae a2 Ia LOT err a SP a oo o— Be ae SPOT ok. AICS eee nee ae ay 43 Hf AE se = —= , Se == Se Fig. 1: a, Puya maculata L. B. Smith, apex of leaf x 1; b, floral bract xX 1; c, flower X 1; d, sepal x 1; e, Puya nutans L. B. Smith, upper scape and inflorescence X 0.5; f, sepal x 1; g, Puya pygmaea L. B. Smith, upper scape and inflorescence x 1; h, flower X 1; i, sepal x 1; j, Pitcairnia Campii L. B. Smith, floral bract and flower x 1; k, sepal x 1. Paramo de Tinajillas and surrounding chaparral and forests, 30-50 km. south of Cuenca, 10,000-10,500 ft., E-2082. Puya maculata L. B. Smith, sp. nov. Fig. 1, a-d. Hefba 5-6 dm. (raro 1 m.) alta (! Camp); foliis rosulatis, ultra 4 dm, longis, crassis et rigidis sed fragilis et facillime fractis (! Camp), vaginis suborbiculari- bus, 4 cm. diametro, pallidis, extus ad apicem versus lepidotis, alibi glabris, laminis lineari-triangularibus, pungentibus, basi 4 cm. latis, supra glabris luci- disque, pallide viridibus sed sub spinis insigniter atro-maculatis, subtus minute albo-lepidotis inter nervos crebros; scapo 12 mm. diametro, brunneo-flocculoso; scapi bracteis dense imbricatis, infimis foliaceis, supremis ovatis, acuminatis, fere integris, ex sicco subpapyraceis, brunneo-lanosis; inflorescentia strobiliformi- ‘clavata, 15 cm. longa, 5 cm, diametro, brunneo-lanosa; bracteis primariis erectis, floribus fere occultantibus, supremis scapi similibus; ramis abortivis, flores 3 fas- ciculatos gerentibus; bracteis florigeris ellipticis, acutis, sepala superantibus, tenuibus; pedicellis obconicis, 6 mm. longis, sepalis oblongo-lanceolatis, sub- acutis, 21 mm. longis, ecarinatis, subcoriaceis; petalis 4 cm. longis, atro-caeru- leis (! Camp). 1952) PLANTS COLLECTED IN ECUADOR 27 Azuay: in open paramo, Paramo del Castillo, crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez, 11,000-11,300 ft., August 21, 1945, Camp E-4882, The curious dark spots at the base of the leaf-spines in Puya maculata con- stitute a character that is unique in the genus. The more technical characters of the species indicate affinity with P. clava-herculis Mez & Sodiro, but it differs from that in the erect primary bracts and subacute rather than acuminate sepals. Puya nutans L. B. Smith, sp. nov. Fig. l,e, f. Herba acaulis, 55 cm. alta; foliis plurimis, densissime rosulatis, vaginis sub- orbicularibus, 35 mm. diametro, extus atro-castaneis, apice adpresse lepidotis, alibi glabris, laminis lineari-triangularibus, pungentibus, 15 cm. longis, basi 12 mm. latis, supra glabris lucidisque, subtus membrana e lepidibus cinereis formata obtectis; scapo ad apicem versus leviter decurvato, 15 mm. diametro, leviter fer- rugineo-lanoso, mox glabro; scapi vaginis erectis, imbricatis, infimis subfoliaceis, alteris latissime ellipticis, serrulatis, ex sicco tenuibus, atris, apice laminis linearibus gradatim brevioribus praeditis; inflorescentia nutans, simplicissima, dense strobiliformi, subglobosa, sub anthesin ca. 5 cm. diametro; bracteis flori- geris eis scapi similibus sed apiculatis et obscure denticulatis, sepala super- antibus; pedicellis ca. 5 mm. longis, robustis; sepalis anguste obovatis, late obtusis, 18 mm, longis, densissime ferrugineo-stellatis; petalis 4 cm. longis, pallide viridibus (! Camp). Azuay: common on Paramo de Tinajillas, 30-50 km. south of Cuenca, 11,000- 11,500 ft., March 17, 1945, Camp E-2291. The nodding inflorescence of Puya nutans appears to be unique in the genus. Discounting the simple inflorescence which is not a reliable character anyway, Puya nutans is probably related to P. clava-herculis Mez & Sodiro. she shape of the sepal is obovate in P. nutans as against triangular in the other. Camp notes: ‘Although thousands of plants were seen, only three inflores- cences were seen during the entire day. This plant is partly injured but rarely killed by the usual paramo fires. Its control must be a real problem; and there is little evidence that there is any success in control short of grubbing. Should the high-altitude paramo pasture ever become valuable, the eradication of this pest will be a real problem.”’ Puya pygmaea L. B. Smith, sp. nov. Fig. 1, g-i. Herba acaulis, 3 dm. alta; foliis multis, rosulatis, ad 17 cm. longis, vaginis parvis, suborbicularibus, serrulatis, dissite lepidotis, laminis lineari-triangulari- bus, pungentibus, basi 15 mm. latis, supra glabris lucidisque, subtus dense ad- presseque albido-lepidotis, spinis hamatis gracilibus brunneis 2.5 mm. longis laxe armatis; scapo gracili sed bracteis inclusis latitudine inflorescentiae fere ae- quante; scapi bracteis densissime imbricatis, ellipticis, lineari-laminatis vel acuminatis, ex sicco papyraceis, valde nervatis, mox glabris; inflorescentia sub- simplici, denssissime strobiliformi, ellipsoidea, 5 cm. longa, 2.5 cm. diametro, albo-lanato; bracteis primariis eis scapi similibus sed minoribus et plus vestitis, sepala multo superantibus, rubris; ramis abortivis cum floribus fasciculatis vel - solitariis; bracteis florigeris ellipticis, acutis, carinatis, membranaceis, sepala superantibus; pedicellis brevibus sed gracilibus; sepalis ellipticis, obtusis, 18 mm. longis, tenuibus; petalis 3 cm. longis, viridi-azureis. Azuay: plants solitary and usually scattered in open paramo, Paramo de Tina- jillas, 30-50 km. south of Cuenca, 11,000—11,500 ft., March 17, 1945, Camp E-2236, The inflorescence of Puya pygmaea is nearly simple but there are floral bracts _ in the axils of the lower primary bracts. The species is so closely allied to Puya 28 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 exigua Mez that I am proposing it as new with considerable misgiving, especially as my knowledge of P. exigua is confined to the description and a photograph, However, the lack of lanate indument on the leaf-apices, the pale color of the indument of the inflorescence and the much larger sepals would indicate that P. pygmaea probably merits specific segregation. At the same time this close rela- tionship indicates an Ecuadorian origin for the heretofore doubtful P. exigua. Pitcairnia aphelandraeflora Lem. Santiago-Zamora (‘‘Oriente’’); plants from heavy rhizomatous base, arching over streams, single non-branched stems, to 3 m., inflorescence bracts crimson tipped with green, becoming green in fruit, perianth crimson; growing only at very margin of swift-running mountain streams where subject to frequent torrents of mountain floods, as Salix in northern hemisphere, along narrow floodplain of Rio Itzintza, Cordillera Cutuct, ca. 2° 40’ S., 78° W., 3,500—3,700 ft., E-1238. The petals are appendaged in the above material while in the type they are naked, but this variation has been recorded in a number of species in Pitcairnia already. The apparently unique character of naked ovules is shown distinctly by both collections. Pitcairnia campii L. B. Smith, sp. nov. Fig. 1, j, k. - Florifera 3 m. alta (! Camp); vaginis foliorum 2 dm. longis, amplis, densissime adpresseque brunneo-lanosis, basi-atro-castaneis, laminis dimorphis, alteris valde reductis, linearibus, 2 mm. latis, spinis uncinatis atris laxe armatis, alteris cum vaginis et petiolo lato canaliculato serrato ad 3 m. longis (! Camp), arcuatis, linearibus, longe acuminatis, 5 cm. latis, basi excepta integris; scapo erecto, basi 2 cm. diametro, pallide flocculoso; scapi bracteis erectis, densissime im- bricatis, ellipticis, pallide flocculosis, infimis longe acuminatis et ad apicem versus spinis atris subdense armatis, supremis acutis, integris; inflorescentia simplicissima, gracile cylindrica, multiflora, 2-3 cm. diametro, petalis exceptis pallide flocculosa; bracteis florigeris erectis vel apice paulo divergentibus, supremis scapi similibus, ad 6 cm. longis sepala bene superantibus; floribus sub- sessilibus; sepalis anguste oblongis, late acutis apiculatisque, 28 mm. longis, ecarinatis; petalis nudis, 7 cm. longis, fulgide aureis (! Camp); staminibus in- clusis, ovario 5/6 supero; ovulis longe caudatis. Junction of Guayas, Cafiar, Chimborazo and Bolivar: on steep banks and cliffs, foothills of the western cordillera near the village of Bucay, 1,000-1,250 ft., June 8—15, 1945, Camp E-3661. In Mez’s key in the Pflanzenreich Pitcairnia campii would fall next to the Mexican P. imbricata, from which it differs in its great size and in its serrate lower*scape-bracts. Pitcairnia heterophylla (Lindl.) Beer. Chimborazo: plants usually leafless at this season (June 19), corolla crimson; on dry rock-faces, cafion of the Rio Chanchan from Naranjapata to below Huigra, 2,000—3,000 ft., E-3888, Pitcairnia pungens HBK. Chimborazo-Cafiar border: on rocky outcrops; lower bracts reddish, upper nitid, green, nigrescent-tipped, calyx yellow, margins and tip red-orange, corolla crim- son, filaments pale crimson; anthers yellow; in Quichua: ‘‘Urcu-huicundo”’ (Urcu== mountain; huicundo—name applied to this general type of plant); westerm escarp- ment, near E] Tambo, 10,000-11,500 ft., E-4085. Cafiar: floral parts deep salmony- scarlet; between Suscal and Chontamarca, north rim of the valley of the Rio de Cafiar, E-2871. Azuay: roots, ground and cooked, used as diuretic; plants branched at base, bracts green, with purple bases, perianth scarlet, anthers yellow, ‘‘Quinde 1952] | PLANTS COLLECTED IN ECUADOR 29 Sungana’’; on rock, along the Rio Matadero, west of Cuenca, 8,500-9,000 ft., E-1940, Perianth segments salmon at base, apically crimson, anthers bright yel- low, ‘‘Quinde-Sangana,’’ root is ground and infusion made which is said to be good for kidneys and liver; Pacific side of pass, Nudo de Portete, pass between head- waters of the Rios Tarqui (Atlantic) and Giron (Pacific), ca. 9,000 ft., E-2172., Subfamily 2. Tillandsioideae Tillandsia complanata Benth. Cafiar: epiphyte, leaves purplish mottled, inflorescences about 5 per leaf- axil, bracts crimson, corolla magenta-pink; uplands called ‘‘Huairacaja,’’ 10-20 km. northeast of Azogues, 11,000 ft., E-1753. Azuay: epiphytic, leaves green, flecked with purple, or blotched, peduncles pale green, upper bracts rosy-pink, corolla lavender-pink, Cruz Pamba region above Bafios (ca. 15 km. southwest of Cuenca), 9,000—10,000 ft., M. Giler & F. Prieto, E-3957A, same, leaves com- pletely reddish-purple on both surfaces, M. Giler & F, Prieto, E-3957B. Tillandsia floribunda HBK. Chimborazo: on dry cliff; inflorescence bracts purplish-red, perianth segments lavender-purple; cafion of the Rio Chanchan near Huigra, 4,000—4,500 ft., (mostly scrub-chaparral, with a few seepages and small swamps along the river), E-3115. On rocks; inflorescence bracts rose-magenta; cafion of the Rio Chanchan, from Naranjapata to below Huigra, 2,000—3,000 ft., E-3860. Tillandsia Hamaleana E. Morr. Chimborazo: leaves purple outside, pale green with purple spots inside, bracts pale green, spotted with purple; on dry rock, cafion of the Rio Chanchan, from Naranjapata to below Huigra, 2,000—3,000 ft., E-3898. Tillandsia latifolia Meyen var. divaricata (Benth.) Mez. Azuay: on cliffs; plants coarse, much-branched, usually loosely attached so that when one part is pulled down a meter or more of stem and attached branches also comes down, leaves silvery, peduncles and bracts red-orange, becoming pale brown in fruit, flowers pale bluish, ‘“‘Guicundo’’; between Rios Azogues and Gualaceo, valley of the Rio Paute, between Paute and Cuenca; 7,200~8,000 ft., (dry cliffs, rocky hillsides and occasional ravines), E-2303. Tillandsia narthecioides Presl. Junction of the Provinces of Guayas, Cafiar, Chimborazo and Bolivar: epiphyte, leaves deep green, peduncle and bracts purple, flowers white, fruit green; foot- hills of the western cordillera near the village of Bucay, 1,000~—1,250 ft., E-3676. Tillandsia pendulispica Mez. Santiago-Zamora (‘‘Oriente’’): Epiphyte, bracts bright red-orange, ridge top south and west of Rio Itzintza, 5,900—6,000 ft., Cordillera Cutuct, ca. 2° 40’ S., 78° W., E-1385. Same locality, plants epiphytic in small clusters, bases radiate from common center, leaf bases and lower bracts reddish-purple, blades bright green, shining, peduncle and its bracts red-orange, outer perianth parts cream yellow, inner sulphur-yellow, E-1392. Tillandsia secunda HBK. Azuay: plant epiphytic, flowering spike 3 m. long with about 40 lateral branches, bracts bright yellow, before anthesis, buds lateral, at anthesis, individ- ual pedicels turn sharply downward, calyces green, stiff, notably fibrous, at anthesis, only 2—3 mm. of corolla exposed, tips of petals flared, lower half of petals white, upper half deep purplish blue, leaves in massive rosette, notably SE - Ce a ee ea ee (6 eee eee ‘ 30 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN (Vol. 8, No. 1 pale green not mottled; the eastern Cordillera, 1—8 km.-north of the village of Sevilla de Oro, 8,000—9,000 ft., E-4592., Tillandsia tetrantha R. & P. var. densiflora (André) L. B. Smith. Chimborazo-Cafiar border: epiphyte, leaves pale green, base inside nigrescent purple, flowering spike and bracts salmon pink, calyx yellow with salmon tint, corolla deep yellow; (western escarpment), near Tipococha, 9,800—10,400 ft., E-4079. Tillandsia tetrantha R. & P. var. scarlatina (André) L. B. Smith. Epiphyte, leaves green above, often with purplish spots, clasping bases usually nigrescent-purple, peduncles and bracts orange-crimson, sepals red- orange below, yellow above, corolla bright yellow; the eastern cordillera, 4—6 km. north of the village of Sevilla de Oro, 9,000—10,000 ft., E-4685. Tillandsia usneoides L. Azuay: sparse, seen hanging from a few plants in the cafion; between Rios Azogues and Gualaceo, valley of the Rio Paute, between Paute and Cuenca, 7,200—8,000 ft., (dry cliffs, rocky hillsides and occasional ravines), E-2348., Loja: hanging from trees, Malacatos Valley, 25 km. south of Loja, about 5,000 ft., E-126. Vriesia arpocalyx (André) L. B. Smith. Azuay: in loose colonies, leaves gray-green, inflorescence spike light pink; on nearly sheer dry cliff, between Rios Azogues and Gualaceo, valley of the Rio Paute, between Paute and Cuenca, 7,200—8,000 ft., E-2345. Vriesia Barclayana (Baker) L. B. Smith. Chimborazo: on rock; leaves pale green, bracts brown-scurfy; cafion of the Rio Chanchan, from Naranjapata to below Huigra, 2,000—3,000 ft., E-3899. Vriesia cylindrica L. B. Smith. Chimborazo: epiphytic, usually in clumps, leaves pale -green, bracts orange- red, sepals yellow, petals green, margins bright blue, tips deep nigrescent-blue, filaments pale greenish-yellow, anthers nigrescent-blue, style nearly white below, nigrescent-blue in twisted zone, stigma greenish; open deforested slope with small patches of scrub in the draws, directly above the village of Huigra, cafion of the Rio Chanchan, 5,000—7,000 ft., E-3493. Azuay: epiphyte, bracts salmony- red, sepals pink-tinged, petals pale green, margins and apices deep (almost nigrescent) blue, leaves pale green, not mottled; the eastern cogdillera, 1—8 km. north of the village of Sevilla de Oro, 8,000—9,000 ft., E-4612. Guzmania lingulata (L.) Mez. Junction of the Provinces of Guayas, Cafiar, Chimborazo and Bolivar: epi- phyte, leaves light green, inflorescence bracts spreading, forming rosette-like head, bright orange-crimson, tips of bracts in head bright yellow; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ft., E-3803. Guzmania minor Mez var. flammea L. B. Smith. Junction of the Provinces of Guayas, Cafiar, Chimborazo and Bolivar: epi- phyte, in small colonies, leaves thin, pale green, subnitid, bracts tipped with deep salmon-pink, basally brownish, later, bracts greenish with. nigrescent-chestnut bases; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ff,, £-3695, Guzmania Pearcei (Baker) L. B. Smith. Santiago-Zamora (‘‘Oriente’’): flowers yellow-greenish, bracts reddish; grow- ing on duff, Cordillera Cutucti, eastern slope and main crest, 6,500 ft., Jorgensen, Cu] -47. 1952] PLANTS COLLECTED IN ECUADOR at Catopsis sessiliflora (R. & P.) Mez. Junction of the Provinces of Guayas, Cafiar, Chimborazo and Bolivar: epi- phyte, leaves, pedicel, and fruit light green; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ft., E-3677. Subfamily 3. Bromelioideae Greigia sodiroana Mez. Azuay: on soil, inflorescences short, leaves lax, spreading, bright green above, inflorescence bracts chestnut-brown; the eastern cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000—9,000 ft., E-4575. The primary bracts in this specimen are very obscurely serrulate but it agrees closely with the following col- lection in all other particulars. Plant on ground, with definite stem, to 0.3 m., inflorescence bracts stiff-coriaceous, basally brown, apically green, corolla bright lavender, individual fruits quite sweet and Giler and Prietomwho have considerable experience in northern Ecuador—say plant is more abundant in north- erm provinces and fruit is sold in Otovalo market under name of ‘‘Pinuela’’; the eastern Cordillera, 4—6 km. north of the village of Sevilla de Oro, 9,000—10,000 ft., E-4704. Aechmea Drakeana Andre. Santiago-Zamora (?): epiphyte, peduncle red, ovary bright pink, perianth bril- liant blue; along the Rio Negro, 4,600 ft., Jorgensen, E-930. Chimborazo: bracts pale green, fruits pink; on soil, moist forested valleys in the afternoon fog-belt, cafion of the Rio Chanchan, about 5 km. north of Huigra, 5,000—6,500 ft., E-3465. CANNACEAE Canna edulis Ker. Azuay: flowers red, leaves used to cover and wrap foods (taking place of ubiquitous banana-leaf of warmer climates), tuberous rhizomes used as food; in Cholo garden, village of Sevilla de Oro, 8,000 ft., E-5029. Canna limbata Rosc. Chimborazo: plants from short tuberous rhizomes, to 4 m. high, perianth parts deep orange-yellow; moist forested valleys in the afternoon fog-belt, Cafion of the Rio Chanchan, about 5 km. north Huigra, 5,000—6,500 ft., E-3432. ' ZINGIBERACEAE Renealmia breviscapa Poepp. & Endl. Chimborazo: plants in clumps, pungent-aromatic, vegetative branches to 2.5 m., leaves dark green above, pale glaucous below, fruits bright red-orange, corolla white, sepals crimson; moist forested valleys in the afternoon fog-belt, cafion of the Rio Chanchan, about 5 km. north of Huigra, 5,000—6,500 ft., E-3382. Renealmia cernua (Sw.) Macbride. Santiago-Zamora (‘‘Oriente’’): plants colonial, to 3 m. high, bracts basally red, apically yellow, tips becoming green with age, perianth yellow, (inflorescence pungent-fragrant); near Mendez, 1,750—2,500 ft., E-896. Renealmia exaltata L. f. Guayas: plants in clumps from short heavy rhizomes, vegetative branches to 5 m., Sheaths yellow-green, leaves deep green above, paler green below, sepals pale pinkish-yellow, corolla-tube pink below, lobes yellow, fruit pale pink; Coastal Plain, in the vicinity of Naranjito, 120 ft., E-3600. 32 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 Renealmia aff. geostachys K. Schum.? Santiago-Zamora (‘‘Oriente’’): plants in colonies, 2.5— zi m. high, inflorescence from rhizome, to 0.3 m., perianth white, pink tinged; ridge ascending into central Cutuct, Cardilter: cueich: ca. 2° 40° S., 78° Way 2.600 ft. o:Badl ee Renealmia thyrsoidea (R. & P.) Poepp. & Endl. Santiago-Zamora (‘‘Oriente’’): plants in large colonies from rhizomatous bases, rhizomes spicy-aromatic, leafy branches to 3 m., inflorescences basal, 1—3 from single base, lower bracts of peduncle red, upper striped red-green, inflorescence bracts clear canary-yellow, perianth parts pale yellow; ridge, western side of Cutuch, Cordillera Cutuct, ca. 2° 40’ S., 78° W., 2,900—3,000 ft., E-1088. Hedychium coronarium Koenig. Guayas: plants in clumps, to 1.5 m. high, rhizomatous, bracts red, leaves pale green, perianth white, large lobes with greenish tinge in center; ciated plain, in the vicinity of Naranjito, ca. 120 m., E-3611. Costus amazonicus (Loes.) Macbr. Santiago-Zamora (‘‘Oriente’’): plants from rhizomatous base, to 1.5 m., leaves dark green above, pale below, bracts green, outer perianth segments pale yellow, inner salmony-red with yellow veins; from Mendez to crossing of Rio Paute, east- er slope of the cordillera, valley of the Rios Negro and Chupianza on the trail from Sevilla de Oro to Mendez, 1,900—2,100 ft., E-1504. Costus argenteus R. & P. Guayas: clumps to G m. high, leaves deep green above, very pale below, bracts basally crimson, apically green, ovary white, sepals bright red with white tips, petals pale yellow, petaloid stamen pale, with lateral salmon-red stripes, center deep yellow, lateral parts with white pubescence internally; coastal plain, in the vicinity of Naranjito, ca. 120 ft., E-3607. Junction of the Provinces of Guayas, Cafiar, Chimborazo and Bolivar: plants in clumps from short rhizomes, to 4 m. high, soft hirsute, hairs brownish, leaves dark green, dull above, paler, nitid below, basal portion of bract crimson, upper half and recurved lobe green, ovary white, subtending bract crimson, seeds black with recurved and crumpled subsucculent tibbonlike appendages; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ft., E-3722. Costus cylindricus Jacq. Junction of Guayas, Cafiar, Chimborazo and Bolivar; plants 1.5—2.5 m. from heavy short-branched rhizomes, lower half of stem with sheaths, upper half with leaves, leaves deep green above, silvery below, bracts deep crimson on both sides €xcept for exposed part externally which is crimson with greenish tinge and with distinct yellow line in center, inner bract crimson, ovary white, sepals crim- son, corolla salmon, sterile stamen salmon-yellow; foothills of the western cor- dillera near the village of Bucay, 1,000—1,250 ft., E-3702. Costus laevis R. & P. Guayas: plants 1 m., leaves pale green, outer bracts crimson externally, in- ternally nigrescent-crimson, inner bract, ovary and sepals deep crimson, petals pale salmon, sterile stamen deep salmon with bright yellow margin and center; coastal plain, in the vicinity of Naranjito, ca. 120 ft., E-3632.: MUSACEAE Heliconia hirsuta L. f. Guayas: plants 1.5—2 m., from short clumped rhizomes, leaves along stem, bracts deep brownish-red, crimson at basal ‘‘joint’’ and at tip, perianth segments 1952] PLANTS COLLECTED IN ECUADOR 33 bright yellow, tips bright green, fruits green; coastal plain, in the vicinity of Naranjito, ca. 120 ft., E-3561. Heliconia latispatha Benth. Guayas: plants in clumps from heavy, much-branched rhizomes, leaves along stem, blades of basal to 1.3 m. long, inflorescence erect, bracts basally yellow, apically and marginally crimson (when expanded, tip green), perianth segments yellow with green margins, immature fruit pale green; coastal plain, in the vicinity of Naranjito, ca. 120 ft., E-3562. Heliconia Schumanniana Loes. Santiago-Zamora (‘‘Oriente’’): plants in clumps, 1.5—2 m., main axis and bract bases deep orange, bract tips and pedicels orange, base of ovary cream-yellow, top of ovary in flower and fruit bright green, perianth canary yellow, leaves green above, glaucous below, but glaucescence wipes off easily; above Rio Upano, near junction with Rio Paute, near Mendez, 1,750—2,500 ft., E-959. Heliconia villosa Kl. Junction of Guayas, Cafiar, Chimborazo and Bolivar: plants in clumps from short-branched rhizomes, ‘‘stems’’ formed by clasping petioles, elliptic in section, about 5 x 10 cm. in diameter, 3—4 leaves per stem, petioles 4—5 m., the free part 1-2 m. long, oval to subterete in section, except near the blade where channeled, blades 2~3 m. long, flowering peduncles breaking out of ‘‘stem’’ about 2.5 m. from ground, usually with single large bract near base, inflorescence pendant, peduncle about 1 m. long, pale red, set with bright golden-orange pubescence, flowering part of inflorescence about 1 m. long, bracts covering flower series, flowers arching downward from vulvular bracts, 25—28 normal bracts plus a some- what larger one at morphological base of inflorescence proper or occasionally this and the very large and distant bract omitted and a medium-sized bract at about the middle of the exposed part of the peduncle, bracts deep orange-crimson, margin and tip shining black, base of perianth white, exposed parts pale yellow, set with golden-yellow hairs, fruit deep lavender, protruding from bract when mature, a regal sight in the jungle; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ft., E-3835. MARANTACEAE Calathea insignis Peters. Junction of the Provinces of Guayas, Cafiar, Chimborazo and Bolivar: plants in clumps from short-branched rhizomes, basal leaves with petioles to 5 m., blades narrowly oval, to 1 m. long, deep green above, paler green below, inflorescence on stem 1—1.5 m. above soil, inflorescences 3, bracts pale yellowish-green, ex- posed parts of perianth pale yellow, seeds nigrescent-blue with large white aril; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ft., E-3683, Calathea lutea (Aubl.) G. F. W. Mey. Guayas: plants in clumps from ascending, woody rhizomes, flowering stems to - 2 m. or more, petioles to 2.5 m., blades to 1 m. long, leaves deep green above, very glaucous below; coastal plain, in the vicinity of Naranjito, ca. 120 ft., i= 599. Calathea macrosepala K. Schum. ; Junction of Guayas, Cafiar, Chimborazo and Bolivar: plants in clumps, from short rhizomes, usually only 1 leaf from basal part of stem, upper surface of leaves bright green, lower pale green but not glaucous, outer bracts pale green, inner ‘ 34. MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 pale green with sterile (?) bracts in axils, apparently usually 2 fertile flowers, outer (or sheathing) perianth segments nigrescent-purple, inner segments (or ster- ile stamens?) salmony yellow, seeds gunpowder-blue, short aril; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ft., E-3783. Calathea nodosa Rusby. Santiago-Zamora (‘‘Oriente’’): plants from rhizomatous base, petioles to 2 m., corolla pale white with purplish tinge, branch accompanying inflorescence droops over later and takes root, axillary branch becoming new plant; Cordillera Cutuct, ca. 2° 40’ S., 78° W., 3,500—3,700 ft., E-1234. Calathea peruviana Koern. Santiago-Zamora (‘‘Oriente’’): plants from short rhizomatous bases, upper sur- face of leaves variegated with three shades of green in perfect scalloped pattern, lower surface with red and green, peduncle reddish, lower (flowering) bracts yellow-green, upper (non-flowering) forming corona-like structure and deep green with red margins on outside, perianth white with purple tinge; Cordillera Cutuca, ca. 2° 40’ S., 78° W., 3,900 ft., E-1108. A variety with nearly glabrous leaves. Calathea picturata K. Koch & Lind. ex char. Junction of Guayas, Cafiar, Chimborazo and Bolivar: in clumps from short off- shoots, leaf dark green with light green zones along center vein and also about half way to margin above, lower surface suffused with purple, bracts purplish, flower parts white; foothills of the westem cordillera near the village of Bucay, 1,000—1,250 ft., E-3844. Myrosma stromanthoides Macbride. Santiago-Zamora (‘‘Oriente’’): plants in colonies from short rhizomes, petioles to 2 m., blades 0.4—0.8 m. long, inflorescence about 1 m. above ground level, bracts and perianth parts yellow-orange, connection between blade and petiole ‘Sointed’’; ridge, western side of Cutuci, Cordillera Cutuct, ca. 2° 40’ S., 78° W., 2,900-3,000 ft., E-1087. ; Pleiostachya Morlaei (Eggers) K. Schum. Guayas: plants in clumps, rhizomatous, petioles of lower leaves to 2 m., blade to 1 m., deep green above and below, veins red below, inflorescence in branched series at several levels, sepals pale yellow, 2 coroila segments yellow, 3rd large, lavender; coastal plain, in the vicinity of Naranjito, ca. 120 ft., E-3594. Vol. 8, No. 1 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN __ March 26, 1952 PLANTS COLLECTED IN ECUADOR BY W. H. CAMP. XYRIDACEAE LYMAN B. SMITH AND JESUS M. IDROBO Xyris acutifolia (Heimerl) Malme. Plants in tufts in mossy swamp, all flower parts bright yellow, bracts deep chestnut brown, small mass of crisp pubescence, yellow, in flower (pubescence of staminodes); F. Prieto P-306. Xyris subulata R. & P. Azuay-“‘Oriente’’ border; plants tufted, flowers yellow; in swamp, Eastern Cordillera, between Ofia and the Rio Yacuambi, 8,000—9,500 ft., F. Prieto P-225, 35 Vol 8, No. 1 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN _ March 26, 1952 PLANTS COLLECTED IN ECUADOR BY W. H. CAMP. BEGONIACEAE LYMAN B. SMITH AND BERNICE G. SCHUBERT Begonia acerifolia HBK. Loja: flowers white, petals frilled, corolla 3 cm. diameter; when immature the large wing of fruit pink, the smaller wings pale green; petioles and pedicels bright salmon color; leaf-blades concolor, dull green (one plant found with variegated leaves); plant rhizomatous, the rhizomes and often the plant base enlarged and tuberous; in moist shaded places, Nudo de Cajanuma, 7 km. south of Loja, 8,000— 8,400 ft., E-121. Begonia aequatorialis Smith & Schubert. Chimborazo: plants on soil (or sometimes epiphytic), from short tuberous rhi- zomes; leaves deep green above, pale below; perianth segments rosy pink; Cafion of the Rio Chanchan, about 5 km. north of Huigra, moist forested valleys in the afternoon fog-belt, 5,000—6,500 ft., E-3268. Begonia aeranthos Smith & Schubert, sp. nov. Fig. 2. Herba e fragmentis solum cognita, alte scandens (! Camp), nodis radicans, omnino glabra; foliis 3-7 mm. supra basin peltatis, rectis vel paulo obliquis, paulo asymmetricis, ovatis, ad 24 cm. longis, 7—11 cm. latis, abrupte acuminatis, basi late truncatis, margine leviter undulatis, minute sparseque denticulatis, Fig. 2: a, Begonia aeranthos Smith & Schubert, leaf x 0.25; b, inflorescence x 0.25; c, staminate flower xX 1; d, stamen X 5; e, style x5; f, Begonia valvata Smith & Schubert, branchlet x 1; g, staminate flower x1; h, young capsule x 1; i, old capsule and peduncle x 1. 36 1952) PLANTS COLLECTED IN ECUADOR aid petiolis 5—20 cm. longis, stipulis mox deciduis, ignotis; pedunculis ad 11 cm. longis, inflorescentiis ample laxeque cymosis, multifloris; bracteis persistentibus, ovatis, acutis, infimis 7 mm. longis; tepalis masculinis 4, rubris, 8—10 mm. longis, obtusis, integris, exterioribus latissime ovatis, interioribus oblongis; staminibus paucis, liberis, antheris oblongis, apice truncatis, quam filamentis paulo longioribus; floribus femineis persenilibus et in parte delapsis solum cog- nitis sed cicatribus indicatis tepalis 5 praeditis, tepalis verisimiliter aequalibus, obovatis, obtusis, 14 mm. longis, integris; stylis 3, basi breviter connatis, pro- funde bifidis, stigmatibus linearibus, spiraliter tortis, placentis integris; capsula subglobosa, 1 cm. diametro, valde inaequaliter alata, ala maxima late ovata, ascendente, 3 cm. longa reliquis marginiformibus; seminibus cylindricis, apice subtruncatis et quadrato-reticulatis, medio reticulis anguste oblongis praeditis. Santiago-Zamora (‘‘Oriente’’): climbing, flowers deep red; ridge ascending into central Cutuct, Cordillera Cutuct, ca. 2° 40’ S., 78° W., 4,400—4,700 ft., Nov. 17-Dec. 5, 1944, Camp E-1143. High climber, leaves dark green above, reddish below, perianth parts and fruit bright crimson; ridge just south and west of Rio Itzintza, Cordillera Cutuct, ca. 2° 40’ S., 78° W., 4,500—5,500 ft., Nov. 17-Dec. 5, 1944, Camp E-1317 (type). At first glance Begonia aeranthos appears to be an oversize specimen of B. glabra, but the peltate leaves and falcate-ascending capsule-wings contradict that identity. Further the linear spiral stigmas of B. aeranthos bar it from section Pritzelia where B. glabra is. In fact, B. aeranthos does not fit any of the sections used by Irmscher in the Pflanzenreich, and requires either the broader definition of Pritzelia or the making of a new section. Begonia albomaculata C. DC. El Oro: plants erect, 0.5 m., perianth segments pink; in Moro-Moro region (about 21 miles west of Portovelo), in dense rain forest, 3,400—4,200 ft., E-618. Begonia buddleiaefolia A. DC. Santiago-Zamora (‘‘Oriente’’): to 0.3 m., flowers red, leaves reddish below; on rocks, between Hacienda Chontal and Santa Elena, 3,400—4,600 ft., eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), E-819. Plants 0.4 m., perianth crimson, stamens and style yellow, wings on fruit crimson, leaves very rugose; ridge between Rios Ontza and Chupiasa, Cordillera Cutuch, ca. 2° 40’ S., 78° W., 4,300—4,700 ft., E-1189, Leaves rugulose, yellow-green above, center pale green below, margin red, flowers greenish-yellow, with red tinges, structures becoming red in fruit; El Partidero, between Rios Paute and Negro, eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), 2,100—3,100 ft., E-1536. Begonia erythrocarpa A. DC. Chimborazo-Cafiar Border (western escarpment): plants scrambling in brush, to 4m. long, often essentially leafless at this season (July 6—9), leaves often with small cornucopia-like processes above vein-junction (leaves taken from non- flowering plants and branches), perianths of both flowers white, ruffled, outer pair of segments often pink-flushed; between Santa Rosa (8,300 ft.) and Joyagshi (9,000 ft.), E-4023. Chimborazo: plants supported in shrubs, on soil, or in fog- forest sometimes epiphytic, plants often much branched near base, with fibrous roots, and branches to 5 m. long, leaves deep green, nitid above, pale below, perianth-segments ruffled, white, outermost with pink flush, ovary with one large wing and 2 marginal and 3 vascular ridges; cafion of the Rio Chanchan, about 5 38 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol: 8, No. I km. north of Huigra, 5,000—6,500 ft., (moist forested valleys in the afternoon fog- belt), E-3435. Cafiar: rhizomatous, plants to 2 m., leaves glossy above, pallid below, two outer segments of perianth of staminate flower white with pink splotch in center, inner two segments pure white, perianth of pistillate flower white; be- tween Suscal and Chontamarca, north rim of the valley of the Rio de Cafar, Manuel Giler, E-2853. Begonia Froebelii A. DC. Chimborazo: plants tuberous, leaves deep green above, pale-pubescent below, peduncles pink, pedicels deep pink to crimson, ovary green with crimson puberu- lence, perianth-segments of both flowers deep pink to crimson, anthers and stig- mas yellow, large plants in shade, small ones on sunny bank, cafion of the Rio Chanchan, about 5 km. north of Huigra, 5,000—6,500 ft., (moist forested valleys in the afternoon fog-belt), E-3289, Cafiar: base tuberous, stem red, floral parts bright crimson, between Tambo and Suscal, 2,000—3,000 meters, north rim of the valley of the Rio de Cafiar, M. Giler, E-2759, Begonia glabra Aubl. Junction of Provinces of Guayas, Cafiar, Chimborazo and Bolivar: plants climbing tree-trunk to 11 m., leaves pale green; foothills of the western cordillera near the village of Bucay, 1,000—1,250 ft., E-3754. El Oro: plants climbing to 5 m. or more, flowers white; in Moro-Moro region (about 21 miles west of Portovelo), 3,400—4,200 ft., in dense rain forest, E-630. Santiago-Zamora (‘‘Oriente’’): to 1.5 m., flowers pink; eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on trail from Sevilla de Oro to Mendez), E-769. Begonia griseocaulis Irmsch. Chimborazo: stems coarse, to 1 m. high, 10 cm. diameter, leafless at this season; peduncles to 1.5 m. long; bracts pink; flowers white to light pink; nearly mature fruit pale greenish, flushed with pink; cafion of the Rio Chanchan, from Naranjapata to below Huigra, 2,000—3,000 ft., E-3862. - Begonia guaduensis HBK. Santiago-Zamora (‘‘Oriente’’): erect, 2 meters, flowers white, 10 km. upstream from Zamora, 4,500 ft., valley of the Rio Zamora, east of Loja, E-54. New to Ecuador. Begonia Holtonis A. DC. (B. Schimpfti Irmscher). Junction of the Provinces of Guayas, Cafiar, Chimborazo and Bolivar: plants much-branched, climbing to several meters, stem and pedicels red, leaves pale green with red margin, flowers flushed with pink; foothills of the westem cor- dillera near the village of Bucay, 1,000—1,250 ft., E-3662. Begonia humilis Ait. Santiago-Zamora (‘‘Oriente’’): low plants on bank; stems red; leaves pale green above, paler below; peduncles bright red; perianths of both male and female flowers white; fruit pale yellow-cream with greenish tint, brown when mature; uplands just south of Rio Chupianza, valley of the Rio Upano, from the Rio Paute north ca. 17 km. to the Chupiangas, 1,950—2,200 ft., E-1442. Begonia Ludwigii Irmsch. Chimborazo: plants from heavy, usually erect stems, sometimes to 1 m. high and 10 cm. diameter at base, inflorescence peduncles to 1 m. long, petioles to 0.5 m. long, leaves pale green, dull above, green or sometimes reddish-tinged below, flowers white, tinged with pink; cafion of the Rio Chanchan, about 5 km. north of Huigra, 5,000—6,500 ft., (moist forested valleys in the afternoon fog- belc), B-3317. . 1952} PLANTS COLLECTED IN ECUADOR 39 Begonia Maurandiae A. DC. Azuay: plants scrambling on bank; the eastern Cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000—9,000 ft., E-4269. Climbing epiphyte; bracts pale green; sepals basally crimson, white band near apex, tip green; ‘‘petals’’ (inner pair) externally crimson, internally pink. Same locality, E-4430. Santiago-Zamora (*‘Oriente’’): epiphyte, leaves dark green above, nitid,; pale and subnitid below, petioles and peduncles red, outer perianth segments reddish or greenish with red veins, inner segments crimson, anthers yellow-green, only staminate flowers seen; Tambo Consuelo to Tambo Cerro Negro, eastern slope of the cordillera, valley of the Rios Negro and Chupianza on the trail from Sevilla de Oro to Men- dez, 8,000—9,000 ft., E-1606. Plant climbing, much-branched; leaves deep green, subnitid above, pale green nitid below; ovary green in flower and immature fruit; mature fruit dead-brown; perianth segments pale green; species apparently dioe- cious (see also no. 4981 for specimen from this region; between Tambo Consuelo and Tambo Cerro Negro, eastern slopes of the cordillera, valley of the Rio Negro, down to the Rio Pailas on the trail to Mendez), 8,500—9,500 ft., Francisco Prieto, E-4974. Begonia maynensis A. DC. Santiago-Zamora (‘‘Oriente’’): to 0.3 m., flowers white; uplands west of Rio Upano, near Mendez, 1,750—2,500 ft., E-977. Begonia octopetala L’Herit. Cafiar: plants from bulbous base, peduncles and petioles crimson, outer peri- anth segments crimson, inner pink to white; near Suscal, north rim of the valley of the Rio de Cafiar, M. Giler, E-2843. Between Suscal and Chontamarca, north rim of the valley of the Rio de Cafiar, M. Giler, E-2891. (inflorescences only, mixed with leaves and stems of another species resembling B. acerifolia). Begonia parviflora Poepp. & Endl. Santiago-Zamora (‘‘Oriente’’): plants to 4 m., flowers white; El Partidero, be- tween Rios Paute and Negro, eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), 2,100—3, 100 ft., E-1528. Plants 2—3 m., leaves seen to 0.75 m. in diameter (these immature), flowers white; Tambo Chontal to Tambo Consuelo, 5,700—8,000 ft., eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Se- villa de Oro to Mendez), E-1592. Begonia piurensis Smith & Schubert. Chimborazo: plants 1—2 m., from heavy tuberous base, stems rooting above tuber, stems red, leaves pale green above, usually reddish below, perianth seg- ments white with red veins, or in some plants deep pink; moist forested valleys in the afternoon fog-belt, cafion of the Rio Chanchan, about 5 km. north of Huigra, 5,000—6,500 ft., E-3294. Same, stem and leaves pale green and flowers pure white, relatively few plants mixed in with number E-3294, probably genetically controlled, E-3294B. New to Ecuador. Begonia Rossmanniae A. DC. Santiago-Zamora (‘‘Oriente’’): plants erect to 1 m., leaves deep green above, paler below, perianth and fruit deep orange red; Yucal region, above the Rio Paute, about 5 km. west of Mendez, 2,400 ft., Jorgensen, E-943. Epiphytic climb- ing vine, leaves deep green above, pale below, flowers cream with greenish tinge, fruit becoming brownish at full maturity; ridge ascending into central Cutuct, 2,600 ft., Cordillera Cutuct, ca. 8° 40'S., 78° W., F. Prieto, E-1132. 1 .) 40 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN (Vol. 8, No. 1 Begonia serotina A. DC. Junction of the Provinces of Guayas, Cafar, Chinmebnan and Bolivar: plants rhizomatous, rhizomes often contorted, ends erect; leaves deep green above, very pale below; pedicels pale greenish; perianth of both flowers white, with pinkish- pale veins; ovary white with pink tinge at anthesis, pale green later; moist cliff bases in shade; foothills of the western cordillera near the village of Bucay; 1,000—1,250 ft., E-3716. Begonia Urticae L. f. Azuay: flowers and fruit crimson; the eastern cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000—9,000 ft., E-4346. To 2 m. leaves deep green above, veins red below, flowers bright crimson; in thicket, same locality; E-4621. Loja—‘“‘Oriente’’ Border: perianth segments bright red; stigmas bright crimson or red; stamens pink; fruit red; crest of the Cordillera de Zamora, east of Loja, ca. 10,000 ft., E-91. With much the same appearance of No. 91, except perianth seg- ments white, fruit green; E-92, Flowers deep red; leaves dark green above, pale below with red veins; plants from thin rhizomes; fruit beaked; wings equal; Nudo de Guagrauma, ca. 12 km. south of Zaraguro, 9,500—10,500 ft., E-136. Begonia valvata Smith & Schubert sp. nov. Fig. 2. Suffruticosa; ramis hirtellis vel glabris; foliis rectis, asymmetricis, hahceo latis, acuminatis, basi dimidiatis, ad 7 cm. longis et 2 cm. latis, serratis, supra atinins et subtus ad nervos birt elite: petiolis 13 mm. longis, hicralite: stipulis tarde deciduis, late ellipticis, 7 mm. longis, integris, apice setiferis, pallide brunneis; pedunculis 30—35 mm. longis; inflorescentiis paucifloris, proterandris; pedicellis masculinis 18 mm. longis; tepalis masculinis 4, subaequalibus, 12 mm. longis, retusis, rubris, exterioribus late ellipticis, juvenilibus insigniter valvatis, interioribus suboblongis; staminibus 4, liberis, antheris oblongis, ob- tusis, quam filamentis longioribus; tepalis femineis perjuvenilibus solum cognitis, verisimiliter 5, quorum uno multo minore quam reliquis; stylis multifidis, seg- mentis linearibus; ovario 3-loculato, placentis bicornutis; capsula late turbinata, apice columna alta praedita, aequaliter 3-cornuta; seminibus crasse ellipsoideis. Santiago-Zamora (‘‘Oriente’’): flowers crimson; between Santa Elena and Tres Ranchos, eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), 2,900—3,300 ft., Nov. 1, 1944, Camp E-825, Leaves deep green above, pale below, veins red, perianth parts orange- crimson; between Tres Ranchos and Chontal, eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Men- dez), 2,700—5,700 ft., Dec. 15, 1944, Camp E-1560 (type). In ‘general habit Begonia valvata closely resembles B. Urticae, but its four stamens indicate closer affinity with B. tetrandra. From this last it differs in its subaequal retuse staminate tepals, elongate anthers, and linear style-branches. We realize that the name valvata does not signalize a particularly distinctive character in Begonia,—it just happens not to have been used before. | Vol. 8, No. 1 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN March 26, 1952 PLANTS COLLECTED IN ECUADOR BY W. H. CAMP VACCINIACEAE ' A. C. SMITH In 1944 and 1945, Dr. W. H. Camp made very extensive collections of plants in Ecuador, among which were included approximately 300 numbers of the family Vacciniaceae. Because of his special studies of this family, Dr. Camp’s material and field-notes are of the greatest interest and value. At his request the writer has undertaken the identification of these specimens and has described those that appear to represent new species. That there are 29 such species described in this paper reflects on the industry and perspicacity of Dr. Camp and his as- sistants, F. Prieto and H. Jorgensen, as well as upon the little-known nature of the flora of Ecuador. In this country and in Colombia the Vacciniaceae reach their highest development in America, both as to number of species and diversity. The first set of the material upon which this paper is based is deposited in the herbarium of the New York Botanical Garden, and a second set in that of the U. S. National Museum. All the specimens of the family collected are cited and are represented in both herbaria unless followed by the annotation ‘‘NY only.’’ A few other specimens are mentioned, the place of deposit being shown by the standard symbols: A (Amold Arboretum); Ch (Chicago Natural History Museum); NY (New York Botanical Garden); US (U. S. National Herbarium). Genera are dis- cussed in the sequence proposed by H. Sleumer’s ‘‘Vaccinioideen-Studien”’ (Bot. Jahrb. 71: 375-408. 1941). Gaylussacia loxensis Sleumer, Bot. Jahrb. 71: 384. 1941. Azuay: ‘‘Oriente’’ Border, east slope of Eastern Cordillera, between Ofia and the Rio Yacuambi, 8,000-9,500 ft. elev., F. Prieto P-265 (NY only) (shrub 1 m.; corolla bright pink). This unicate collection agrees very well with the original description of G. loxensis, but I have not compared it directly with type material (Lehmann 4965 pro parte). As compared with the description, our specimen differs only in its somewhat more pilose flowers and slightly longer corolla (to 9 mm. long at anthesis). Themistoclesia inflata A. C. Smith, sp. nov. Frutex epiphyticus pendulus, ramulis gracillimis elongatis cinereis juventute obscure puberulis mox glabratis; petiolis subteretibus incrassatis rugulosis 2-2.5 mm. longis; laminis in vivo succulentibus inflatisque in sicco coriaceis fuscis, ovato-ellipticis, 3.5-6 cm. longis, 1.5-3.2 cm. latis, ad basim truncato- rotundatum vel minute auriculatum gradatim angustatis, apice in acuminem circiter 1 cm. longum acutum terminantibus, margine incrassatis et leviter recurvatis, supra glabris, subtus pilos breves castaneos glandulosos in foveolis minutis depressos gerentibus, e basi ut videtur 5-nerviis, costa et nervis secundariis supra inconspicue impressis vel ut venulis immersis; floribus supra-axillaribus | ut videtur solitariis vel paucis in rhachi minuta subfasciculatis, bracteis oblongis acutis haud 1 mm. longis hispidulo-marginatis; pedicellis gracilibus sub anthesi 12-17 mm. longis leviter curvatis pilis albidis 0.2-0.3 mm. longis copiose hispidulo- puberulis basim versus obscure bibracteolatis, bracteolis subulatis haud 0.5 mm. longis caducis; calyce turbinato sub anthesi 5-6 mm. longo et circiter 4 mm. apice ’ diametro ut pedicellis puberulo, tubo inconspicue 5-angulato 3-4 mm. longo, limbo suberecto quam tubo breviore intus glabro inconspicue 5-lobato, lobis late 41 ik 42 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol. 8, No. 1 deltoideis apiculatis 0.51 mm. longis, sinibus rotundatis; disco annulari-pulvinato glabro; corolla glabra tenuiter camosa late urceolata circiter 7 mm. longa et medium versus 4.5 mm. diametro, basi et apice paullo angustata, lobis 5 deltoideis subacutis circiter 1.5 x 2 mm. sub anthesi reflexis; staminibus 10 longitudine corollam fere aequantibus leviter inaequalibus, filamentis pallidis ligulatis liberis alternatim circiter 1.5 mm. et 1.8 mm. longis superne pilis circiter 0.5 mm. longis dorso hispidulis, antheris 4.5=5 mm. longis, thecis granulatis 1.7-2 mm. longis basi obscure mucronulatis, tubulis gracillimis quam thecis longioribus et multo angustioribus per rimas circiter 0.5 mm. longas dehiscentibus; stylo tereti corollam subaequante basim versus leviter incrassato, stigmate minuto. Chimborazo: Cafion of the Rio Chanchan, about 5 km. north of Huigra, 5,000- 6,500 ft. elev. (moist forested valleys in the afternoon fog-belt), May 19-28, 1945, Camp E-3363 (TYPE US 1,989,082; dupl. NY) (hanging epiphyte; leaves crisp- succulent, to 4 mm. thick, dark translucent-green, inflated above, pale below with stomata, the surface with numerous pits, the bases of glands; lower % of the leaf in cross-section with chlorenchyma, the upper % with apparently 4 cell-layers of hydrenchyma; corolla cream-white, the anthers orange). This very distinct species is most closely allied to the Colombian T. rostrata A. C. Smith, from which it differs in having its leaf-blades somewhat smaller, narrower at base, thicker and inflated when fresh, and with immersed venation and gland-hairs sunk in small pits on the lower surface. The new species is further distinguished by its very slender anther-tubules, which are not charac- teristic of Themistoclesia. Slender tubules of this type have been noted in the Colombian T. crassifolia Sleumer (Bot. Jahrb. 71: 392. 1941), to which T. inflata is also allied. However, T. crassifolia has much larger and obviously cordate leaf-blades and an elongate inflorescence, among other characters which dis- tinguish it from the new species. Themistoclesia cutucuensis A. C. Smith, sp. nov. Frutex epiphyticus, ramulis gracilibus fuscis obtuse angulatis juventute albido- puberulis mox glabratis; stipulis intrapetiolaribus subulatis 3-5 mm. longis primo puberulis; petiolis leviter canaliculatis 2-5 mm. longis ut ramulis pube- rulis; laminis chartaceis vel subcoriaceis in sicco pallide viridibus vel metallico- olivaceis, anguste elliptico-oblongis, (4—)6=10 cm. longis, (1.2-)1.5-2.5 cm. latis, basi anguste rotundatis vel subcordatis, apice angustato obtusis, margine integris valde revolutis, juvenilibus utrinque parce strigillosis (pilis brevibus castaneis glandulosis) etiam puberulis, supra mox subtus demum glabratis, costa supra impressa subtus elevata, nervis secundariis principalibus utrinsecus 2 vel 3 incongpicuis e costa basim versus orientibus utrinque prominulis, rete venularum utrinque prominulo vel subimmerso; inflorescentia axillari solitaria racemosa 8—12-flora bracteis paucis papyraceis lanceolatis 3-5 mm. longis parce puberulis basi circumdata, rhachi gracili angulata sub anthesi pedunculo 1-2 cm. longo incluso 5=11 cm. longa albido-puberula, bracteis sub floribus acuminatis ut eis basi inflorescentiae; pedicellis gracilibus sub anthesi 7-12 mm. longis ut rhachi puberulis, superne parce glanduloso-strigillosis, medium versus bibracteolatis, bracteolis ut bracteis sed 1.5-2 mm. longis; calyce sub anthesi 5-6 mm. longo et apice circiter 3 mm. diametro parce puberulo, tubo 5-angulato 3=3.5 mm. Jongo parce glanduloso-strigilloso, limbo erecto submembranaceo’ profunde 5-lobato, lobis elongato-deltoideis 1.5=2.5 mm. longis acutis, sinibus acutis; disco annulari glabro; corolla submembranacea urceolato-cylindrica sub anthesi 7-9 mm. longa et basim versus alis inclusis circiter 3 mm. diametro, supeme angustata et pallide puberula, 5-alata, alis basim versus 0.6-1 mm. latis supeme angustioribus, lobis deltoideo-lanceolatis circiter 2 mm. longis; staminibus 10, filamentis incon- 1952] PLANTS COLLECTED IN ECUADOR 43 spicuis liberis ligulatis glabris circiter 1 mm. longis, antheris 6.5-7 mm. longis, thecis 1.5-1.8 mm. longis basi obscure mucronulatis, tubulis quam thecis triplo longioribus per rimas elongatas dehiscentibus; stylo filiformi corollam subaequante, stigmate minuto. Santiago-Zamora: Cordillera Cutucu, east-trending slope from top of ridge down toward the Itzintza, 4,800-5,800 ft. elev., Nov. 17-Dec.5, 1944, Camp E-1366 (TYPE US 1,989,018; dupl. NY) (high-growing epiphyte, a semi-vine; leaves dark green above, pale beneath). Same locality, ridge just south and west of Rio Itzintza, 4,500 to 5,900 ft. elev., Camp E-1330, E-1342 (high-growing epiphytes, brought down by storm; leaves shining on both sides). The closest relative of the new species appears to be T. schultzeae Sleumer (Bot. Jahrb. 71: 393. 1941), which also occurs in the ‘“‘Oriente’”’ of Ecuador. From this, T. cutucuensis differs in its somewhat larger leaves and inflorescences, and especially in its larger calyx with elongate lobes and sharp sinuses, its winged corolla, and its differently proportioned anthers, of which the thecae are much shorter and the tubules proportionately longer. The occurrence of a winged corolla in Themistoclesia has otherwise been noted in the Colombian T. pterota A. C. Smith (Jour. Arnold Arb. 27: 103. 1946), which has broader and deeply cordate leaf-blades, a minutely denticulate calyx-limb, and a more strongly urceolate corolla, Themistoclesia campii A. C. Smith, sp. nov. Frutex ad 1.5 m. altus, ramulis gracilibus subteretibus fusco-cinereis albido- puberulis vel fusco-hispidulis ac etiam squamis parvis irregularibus nigrescentibus leprosis, demum glabratis; petiolis subteretibus 2-3 mm. longis primo minute hispidulis demum glabrescentibus; laminis subcoriaceis in sicco fuscis ovato- ellipticis, 2-3.7 cm. longis, 1-2 cm. latis, basi rotundatis vel subcordatis, ad apicem breviter acuminatum callosum gradatim angustatis, margine leviter in- crassatis et recurvatis, supra pallide puberulis mox glabratis, subtus squamis parvis nigris leprosis etiam praecipue costa parce fusco-hispidulis, costa supra leviter impressa subtus valde elevata, nervis lateralibus utrinsecus plerumque 2 e basi orientibus supra leviter impressis vel utrinque ut rete venularum immersis; inflorescentia axillari breviter racemosa 3-10-flora bracteis pluribus papyraceis ovatis subacutis 1-2 mm. longis basi circumdata, rhachi gracili 5-10 mm. longa pilis pallidis 0.2-0.3 mm. longis minute hispidula, bracteis sub floribus ut eis basi inflorescentiae; pedicellis gracilibus sub anthesi 7-11 mm. longis ut rhachi hispidulis etiam nigro-leprosis, medium versus bibracteolatis, bracteolis circiter 1.5 mm. longis; calyce sub anthesi circiter 5 mm. longo et apice 4 mm. diametro ut pedicello hispidulo et leproso, tubo circiter 3 mm. longo manifeste 5-angulato, limbo suberecto papyraceo intus glabro quam tubo breviore 5-lobato, lobis late deltoideis cuspidatis circiter 1 mm. longis, sinibus obtusis; disco annulari-pulvinato hispidulo; corolla carnosa subcylindrica sub anthesi 7.5-8 mm. longa, basim versus 3.5-4 mm. diametro supeme angustata, extus pilis stramineis circiter 0.5 mm. longis hispidula, lobis 5 deltoideis subacutis 1-1.5 mm. longis; staminibus 10 simili- bus quam corolla brevioribus, filamentis liberis gracilibus pallidis ligulatis . circiter 2 mm. longis pilis paucis circiter 0.5 mm. longis hispidulis, antheris 4,5-5.5 mm. longis, thecis 1.5=2 mm. longis basi subacutis, tubulis quam thecis longioribus per rimas circiter 1 mm. longas dehiscentibus; stylo tereti corollam subaequante, stigmate minuto. Azuay: ‘‘Oriente’’ Border, Paramo del Castillo and surrounding forested areas (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), 11,000-11,350 ft. elev., Dec. 17, 1944, Camp E-1640 (TYPE US 1.989,027; dupl. NY) (east of El Pan; shrub 0.5 m. high, terrestrial; leaves deep green above, pale ‘ 44 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol. 8, No. 1 beneath; hypanthium green; corolla salmon-pink in bud, becoming red at anthesis). Same locality, Camp E-4865 (NY only) (shrub to 1.5 m.; leaves crisp-coriaceous, deep green, nitid above, pale beneath, cusped, even the youngest with a black scurf; corolla dull crimson, the lobes yellow). Themistoclesia campii is most readily characterized by the blackish scurfy scales which persist on the lower surfaces of leaves and on the inflorescence, and by its hispidulous corolla. From the Colombian T. compacta A. C. Smith it is distinguished not only by the different character of its foliar indument, but also by the shorter and sparser pubescence of its inflorescence (especially the corolla), its longer pedicels, and its isomorphic stamens with essentially glabrous fila- ments. From T. dependens (Benth.) A. C. Smith the new specie s is readily dis- tinguished by its indument and its short stamens. Not included in the above description, but almost certainly referable to this species, is Camp E-751 (NY only) (Azuay: same locality, 9,000-11,000 ft. elev.; spreading shrub with deep red flowers), which differs only as follows: corolla at anthesis 10-11 mm. long and hispidulous only at apex; filaments about 2.5 mm. long, glabrous; anthers about 7 mm. long, the thecae about 2.5 mm. long. Except for the slightly larger and essentially glabrous corolla and stamens, this specimen agrees perfectly with the two described; it apparently represents a form of the species occurring at somewhat lower elevation. When the full range of variability is known, no. 751 will doubtless fit into a reasonable species concept. Themistoclesia dependens (Benth.) A. C. Smith, Contr. U. S. Nat. Herb. 28: 442. 1932. Azuay: ‘‘Oriente’’ Border, Eastern Cordillera. between Ofia and the Rio Yacu- ambi, east slope, 8,000-9,500 ft. elev., F. Prieto P-270 (shrub to 3 m.; young leaves pale, the old leaves of previous season deep green and subnitid above, pale green and dull beneath; corolla bright crimson, with pale pink lobes). A fairly frequent species in western Colombia and Ecuador, also represented from Azuay by Steyermark 53446 (A, Ch). Sphyrospermum buxifolium Poepp. & Endl. Nov. Gen. & Sp. 1: 4. pl. 8. 1835. Guayas, Cafiar, Chimborazo, & Bolivar junction: Foothills of the western cor- dillera near the village of Bucay, Camp E-3833 (NY only). El Oro: In Moro-Moro region, about 21 miles west of Portovelo, Camp E-632, E-634 (both NY only). Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, near El Topo, Camp E-1686, E-1688 (NY only), E-1689, E-1691 (NY only); same general region, east of Puyo, Camp E-1698 (NY only). Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza, El Partidero, between the Rios Paute and Negro, Camp E-1520 (NY only). Low hills west of Rio Upano, along Rio Chupiangas, F. Prieto ChuP-18. Cordillera Cutuct, Camp E-1181, E-1202, E-1341, E-1361 (all NY only). Field notes accompanying this excellent suite of specimens indicate that the plant is a slender epiphyte, often pendant and high-climbing, rarely scrambling or trailing along banks, and in.one case (no. 634) terrestrial and suberect to 50 cm. high; it occurs in forest at elevations of 2,100 to 5,900 ft.; the corolla is white and sometimes pink-tinged; the mature fruit is light blue or pale bluish lavender, sometimes subtranslucent, and insipid. The related S. majus Griseb. was mentioned in my revision (Brittonia 1: 209. 1933) as having a southern limit of Venezuela, but I have since seen several speci- mens from Colombian Departments as far south as El Valle and Cundinamarca. However, at least in its typical form, the species does not seem to occur in Ec- uador, although some of the cited specimens (e. g. those from the Cordillera 1952] , PLANTS COLLECTED IN ECUADOR 45 Cutuci) approach it in their comparatively large ovate leaves. The value of S. majus as a specific entity at any rate is open to question, and | think it advisable to refer all the small-leaved Ecuadorian specimens with 4 or 5 stamens to S. buxifolium. Sphyrospermum flaviflorum A. C. Smith, sp. nov. Frutex epiphyticus, ramulis gracilibus teretibus cinereis ad nodos incrassatis juventute - albido-puberulis mox glabratis; foliis congestis, petiolis subteretibus 1-2 mm. longis ut ramulis mox glabratis; laminis in sicco coriaceis fusco-olivaceis ellipticis, 15-22 mm. longis, 10-15 mm. latis, basi et apice rotundatis, margine incrassatis paullo recurvatis, subtus minutissime et subpersistenter albido- puberulis ac etiam dispersim glanduloso-strigillosis (pilis castaneis circiter 0.2 mm. longis caducis), e basi obscure 3- vel 5-nerviis, costa supra leviter impressa subtus elevata, nervis lateralibus obscuris; floribus axillaribus solitariis basi bracteis pluribus minutis imbricatis suffultis, rhachi subnulla; pedicellis gracili- bus 2.5-3.5 mm. longis pilis 0.3-0.6 mm. longis albido-hispidulis basi minute bibracteolatis cum calyce continuis; calyce sub anthesi 3.5-4 mm. longo et apice diametro ut pedicellis hispidulo, tubo cupuliformi circiter 2 mm. longo, limbo erecto-patente profunde 5-lobato, lobis elongato-deltoideis circiter 1.5 mm. longis acutis intus glabris, sinibus obtusis vel rotundatis; disco annulari-pulvinato glabro; corolla subcarnosa extus parce hispidula campanulata, sub anthesi circiter 5.5 mm. longa et apice 4 mm. diametro, lobis 5 late deltoideis circiter 0.5 mm. longis subacutis; staminibus 10 alternatim leviter inaequalibus quam corolla brevioribus, filamentis gracilibus ligulatis glabris circiter 1.5 mm. longis, antheris 2.7-3 mm. longis, thecis 1-1.2 mm. longis, tubulis longioribus per poros subterminales cir- citer 0.2.mm. longos dehiscentibus; stylo filiformi corollam subaequante, stig- mate minuto. Cafiar: Valley of Rio de Cafiar, near Rosario, 3,400 ft. elev., Sept. 6-10, 1944, F. Prieto CP-16 (TYPE US 1,988,906; dupl. NY) (epiphyte; leaves succu- lent; corolla yellowish, tinged with pink if exposed to sun). Sphyrospermum f{laviflorum belongs to a small group of Ecuadorian species not discussed in my revision of 1933, but treated in Sleumer’s more recent key (Bot. Jahrb. 71: 394, 395. 1941). The two species thus far comprising the group are characterized by having stamens twice as many as corolla-lobes, very short pedicels, small flowers, and anthers with slender tubules dehiscing by subterminal pores. The new species differs from S$. spruceanum Sleumer in its larger and obscurely pilose leaves and its obviously pedicellate flowers, which are larger in all parts. From the related S. microphyllum Sleumer, it is distinguished by its much larger leaves and flowers, its comparatively conspicuous calyx-lobes, and its campan- ulate (rather than apically contracted) corollas. Sphyrospermum campii A. C. Smith, sp. nov. Frutex epiphyticus vel terrestris, ramulis gracillimis brunneis minute et subper- sistenter pallido-puberulis; petiolis gracilibus teretibus 1-2 mm. longis ut ramulis puberulis; laminis in sicco subcoriaceis fuscis ovatis, 15-26 mm. longis, 10-15 mm. latis, basi late obtusis vel subrotundatis, apice subacutis, margine leviter © recurvatis, subtus evanescenter puberulis ac etiam dispersim glanduloso-strigillo- Sis, e basi obscure 3- vel 5-nerviis, costa supra plana vel paullo impressa subtus subelevata, nervis aliis immersis; floribus axillaribus solitariis; pedicellis graci- libus sub anthesi 1.5-2 mm. longis parce puberulis cum calyce continuis basi bracteis paucis minutis suffultis, basim versus bibracteolatis, bracteolis reniformi- orbiculatis haud 1 mm, latis; calyce sub anthesi circiter 3 mm. longo et apice diametro pilis pallidis 0.3-0.5 mm. longis copiose hispidulo etiam obscure castaneo- glanduloso, tubo subgloboso circiter 1.5 mm. longo, limbo papyraceo suberecto ‘ 46 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol. 8, No. 1 4-lobato intus glabro, lobis late deltoideis 0.5-1 mm. longis subacutis, sinibus rotundatis vel late obtusis; disco annulari-pulvinato glabro; corolla subcarnosa late campanulata sub anthesi 8-9.5 mm. longa, basi angustata, apice circiter 6 mm. diametro, glabra, lobis 4 deltoideis subacutis circiter 1.5 mm. longis et 3-4 mm. latis; staminibus 8 similibus quam corolla brevioribus, filamentis ligu- latis 2—2.5 mm. longis superne copiose pallido-hispidulis, antheris 3.5-3.8 mm. longis, thecis crassis basi obtusis longitudiné tubulos aequantibus, tubulis gracilibus per poros subterminales dehiscentibus; stylo filiformi quam corolla breviore, stigmate truncato. El Oro: In Moro-Moro region, about 21 miles west of Portovelo, 3,400-4,200 ft. elev., Oct. 7, 1944, Camp E-633 (TYPE US 1,988,987; dupl. NY) (in dense rain-forest, seen both as epiphyte and on soil on steep banks; when on banks, the flowering branches ascending to 10-15 cm.; corolla deep red; fruit angled, green- ish white, translucent when mature, 0.75-1 cm. in diameter). The species here described seems best placed as a relative of the three species discussed immediately above, differing from all of them in its ovate and subacute leaf-blades and large corollas. In leaf-size it most closely approximates © S. flaviflorum, described above, but the 4-merous flowers, comparatively small calyx-lobes, glabrous corollas, and differently proportioned anthers further dis- tinguish it. The mention in Dr. Camp’s field notes of an angled fruit may suggest that the new species would be better placed in Themistoclesia, in which it bears a super- ficial resemblance to T. cuatrecasasii A. C. Sm. However, the calyx-tube in flower appears essentially globose, as in Sphyrospermum, and the habit of the plant certainly suggests this genus. The campanulate corollas and nearly terminal anther-pores are known to occur in Sphyrospermum, but not in Themistoclesia. A close approach of the two genera, however, is here indicated. Sphyrospermum cordifolium Benth. Pl. Hartw. 222, 1846, Pichincha: Along the road from Quito to Sto. Domingo de los Coloraiee: Camp E-1734, Azuay: The eastern Cordillera, vicinity of the village of Sevilla de Oro and 1-8 km. northward, Camp E-4442 (NY only), E-4560, E-4740, s. n. (July-Sept. 1945) (NY only). El Oro: In Moro-Moro region, about 21 miles west of Portovelo, Camp E-631 (NY only), Loja: ‘‘Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, Camp E-7/7. Hda. Anganuma, at headwaters of Rio Cachiyacu, on west slopes of Cordillera Condor, about 46 km. south of Loja, Jorgensen & Prieto P]-50B (NY only). Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza, region of Tambo Consuelo, Camp E-1602 (NY only). The cited specimens were collected at elevations of 3,400 to 10,000 ft.; the plant is usually an epiphytic vine, but sometimes terrestrial with branches hang- ing over rocks; leaves crisp-coriaceous, deep green above and paler beneath; corolla white to pinkish or crimson (in plants not otherwise distinguishable); fruit pale blue, translucent. These collections are fairly diverse, but I expect that the range of variation which I previously indicated (Brittonia 1: 213, 214. 1933) is not exceeded; degree of floral pubescence and size of corolla seem to be unstable in this species, and flower-color also seems unreliable. The most extreme of the cited specimens is no. 1734, which perhaps ought to be excluded from the species; it is a shrub 1 m. high, with very short (5-7 mm.) pedicels and small, essentially glabrous flowers. Such short pedicels are found in S. sodiroi (Hoer.) A. C. Smith, also from the Province of Pichincha, but in that species the flowers are densely villose, with comparatively large corollas and long filaments. Sphyrospermum sodiroi is re- corded (Haught 3228a, US) as abundant along the Quito-Santo Domingo road, the 1952) PLANTS COLLECTED IN ECUADOR AT precise locality of Camp’s no. 1734, indicating the possibility that the two species hybridize in this region, if, indeed, S. sodiroi is more than a very extreme form of S. cordifolium. Sphyrospermum sp. Santiago-Zamora: Valley of the Rio Zamora, east of Loja, near Zamora, about 3,000 ft. elev., Camp E-16 (epiphyte, over river; fruit pale blue), This fruiting specimen represents a species not otherwise in Dr. Camp’s mate- rial. In the texture and general shape of its leaves it suggests the Peruvian S, weberbaueri (Hoer.) A. C. Smith, but it differs in the leaf-blades with rounded- obtuse apices and in the short-pedicellate fruits. I think it probable that the specimen represents an undescribed species. Eleutherostemon Herzog. A note on this little genus and a discussion of the eight species now known to compose it have recently been published by the writer (Contr. U. S. Nat. Herb. 29: 350-355. 1950). Collections of Eleutherostemon are still rare, and conse- guently the eight numbers of it obtained by Dr. Camp are very welcome. The only previous records of the genus in Ecuador refer to E. amplectens (Sleumer) A. C. Smith and E. octandrum (Sleumer) A. C. Smith (as Diogenesia octandra Sleumer, Bot. Jahrb. 71: 396. 1941), and it is possible that the latter record is referable to one of the species I describe below as new. Eleutherostemon amplectens is not among Dr. Camp’s collections, all of which, in my opinion, represent undescribed species. Eleutherostemon floribundum A, C. Smith, sp. nov. Frutex interdum epiphyticus multiramosus, ramulis fusco-cinereis obtuse angulatis juventute minute puberulis mox glabratis; stipulis inconspicuis circiter 1 mm. longis basi pulvinatis acutis; petiolis semiteretibus 1.5-3 mm. longis supra subpuberulis; laminis papyraceis in sicco fuscis lanceolatis, 5-10 cm. longis, 1.3-3 cm, latis, basi acutis vel obtusis, in apicem gracilem ad 15 mm. longum mu- cronulatum gradatim angustatis, margine leviter recurvatis, supra glabris, subtus minute et dispersim glanduloso-strigillosis vel eglandulosis, costa supra insculpta subtus valde elevata, nervis secundariis utrinsecus plerumque 2 adscendentibus inconspicuis subtus leviter prominulis, rete venularum immerso vel subtus haud prominulo; inflorescentia axillari breviter racemosa 4-8-flora bracteis numerosis ovatis acutis 1-2 mm. longis basi circumdata, rhachi gracili 3-5 mm. longa minute puberula mox glabrata, bracteis floriferis eis basi rhachis similibus caducis; pedicellis gracilibus sub anthesi et fructu juvenili 9-13 mm. longis ut rhachi minute puberulis et superne parce glandulosis medium versus bibracteolatis, bracteolis lanceolatis 1-1.5 mm. longis caducis; calyce cupuliformi sub anthesi 2.5-3 mm. longo et diametro obscure puberulo glabrato, limbo subcoriaceo erecto minute 4- denticulato (dentibus haud 0.3 mm. longis), sinibus complanatis; disco conspicuo carnoso cylindrico circiter 0.7 mm. alto glabro; corolla tenuiter carnosa cylindrico- subcampanulata, sub anthesi 5-6 mm. longa glabra conspicue 4-lobata, lobis elongato-deltoideis circiter 2 mm. longis et 1.5 mm, latis acutis reflexis distaliter | intus obscure papillosis; staminibus 5 corollam subaequantibus, filamentis ligu- latis liberis 3-3.5 mm. longis ubique copiose breviter pilosis, antheris circiter 3 mm. longis basi obtusis, thecis quam tubulis per rimas 0.6-0.8 mm. longas de- hiscentibus paullo longioribus; stylo crasso tereti leviter exserto basi disco arcte cincto, stigmate minuto; fructibus immaturis ellipsoideis leviter quadrangulatis calycis limbo et disco persistentibus coronatis. Azuay: ‘‘Oriente’’ Border, Paramo del Castillo and surrounding forested areas (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), ‘ 48 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. I 9,000 -11,000 ft. elev., Oct. 30, 1944, Camp E-706 (TYPE US 1,988,992; dupl. NY) (shrub 1.5 m.; flowers white). Santiago-Zamora: Eastern slope of the cor- dillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), between Hda. Chontal and Sta. Elena, 3,400-4,600 ft. elev., Camp E-792 (epiphytic shrub to 3 m.); same locality, between Tambo Chontal and Tambo Consuelo, 5,700-8,000 ft. elev., Camp E-1572 (much-branched epiphyte, in clumps; leaves dark above, paler beneath, shining; fruit globular, pale). In foliage and in calycine characters E. floribundum suggests E. octandrum (Sleumer) A. C. Smith, from which it differs in its more compact and fewer-flowered inflorescences, conspicuous cylindric disk (scarcely 0.2 mm. high in E. octandrum), slightly shorter corolla, and reduced number of stamens, with longer and copiously pilose filaments. From E. tetrandrum A. C. Smith, which similarly has a reduced number of stamens and a conspicuous disk, the new species differs in its propor- tionately narrower leaf-blades, its inflorescence with a shorter rachis but longer pedicels, shorter and more deeply lobed corolla, and stamens with shorter fila- ments and longer anthers. So far as observed at present, the stamens are uniformly 4 in E. tetrandrum and 5 in E. floribundum. Eleutherostemon oliganthum A. C. Smith, sp. nov. Frutex interdum epiphyticus, ramulis gracilibus obtuse angulatis juventute minute puberulis mox glabratis; stipulis obscuris pulvinatis; petiolis subteretibus rugulosis 1+2 mm. longis obscure puberulis glabrescentibus; laminis papyraceis in sicco fuscis ovato-lanceolatis, (4-)5-12 cm. longis, (1.5-)1.8-4 cm. latis, basi late obtusis vel rotundatis vel inconspicue subcordatis, in acuminem ad 2 cm. longum (apice ipso obtuso) angustatis, margine anguste recurvatis, utrinque glabris (juvenilibus basim versus puberulis) vel subtus dispersim et sparsissime glanduloso- strigillosis, costa superne impressa subtus elevata, nervis secundariis utrinsecus 2 vel 3 basim versus orientibus (intimis cum costa ad 2 cm. concurrentibus) adscendentibus supra insculptis vel prominulis subtus paullo elevatis, rete venu- larum supra subplano subtus prominulo; floribus axillaribus solitariis, bracteis basalibus minutis; pedicellis gracillimis parce puberulis sub anthesi 10-13 mm. longis basim versus minute bibracteolatis; calyce sub anthesi 2.5-3.5 mm. longo et 2-3 mm. diametro pilis pallidis circiter 0.2 mm. longis copiose hispidulo ac etiam obscure rubro-glanduloso demum subglabrescente, tubo cupuliformi 1.5-2 mm. longo, limbo subpatente quam tubo leviter breviore minute 5-dentato (dentibus circiter 0.5 mm. longis), sinibus rotundatis; disco carnoso annulari-pulvinato glabro; corolla tenuiter carnosa cylindrica sub anthesi 10-11 mm. longa et circiter 3.5 mm. diametro glabra, lobis 5 deltoideis obtusis haud 1 mm. longis; staminibus 10 quam corolla paullo brevioribus, filamentis gracilibus ligulatis 5-6 mm. longis superne parce hispidulis, antheris 3.5-4 mm. longis, thecis 1.2-1.4 mm. longis basi ro- tundatis vel obscure mucronulatis, tubulis quam thecis longioribus per rimas elongatas dehiscentibus; stylo leviter exserto filiformi, stigmate minute peltato. Santiago-Zamora: Cordillera Cutuct,, along narrow flood-plain of Rio Itzintza, 3,500-3,700 ft. elev., Nov. 17-Dec. 5, 1944, Camp E-1230 (TYPE NY) (epiphytic; leaves shining; corolla crimson); ridge ascending into central Cutuct, 4,400~4, 700 ft. elev., Camp E-1158 (NY only) (shrub; leaves shining beneath; corolla red); on banks of Rio Itzintza, 3,500 ft. elev., Camp E-1205 (NY only) (epiphyte, flowering on old wood among the root's; leaves deep green and dull above, pale and shining beneath; corolla bright crimson). Eleutherostemon oliganthum is characterized by its solitary flowers and elongate filaments, being readily distinguished from E. octandrum (Sleumer) A. C. Smith, apparently its closest relative, by these characters and by its pilose calyx, longer corolla,and more numerous stamens. 1952] | PLANTS COLLECTED IN ECUADOR A9 The following specimen probably also belongs here: Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, on cliff-top, near El Topo, 4,400 ft. elev., Camp E-1687 (plants scrambling from mossy bank, or sometimes epiphytic; fruit solitary, immature, subgobose, about 8 mm. in diameter). It agrees with E. oliganthum in length of pedicel, pubescence of calyx, and the comparatively inconspicuous disk; however, its leaves are slightly different in shape and venation, and I hesitate to make a positive identification of it without flowers. Eleutherostemon gracilipes A. C. Smith, sp. nov. Frutex praeter flores ubique glaber, ramulis gracilibus obtuse angulatis; stipulis obscuris coriaceis pulvinatis haud 1 mm. longis; petiolis subteretibus rugulosis 1-2 mm. longis; laminis papyraceis in sicco fuscis lanceolatis vel ovato-lanceolatis, 7-11 cm. longis, 1.5-5 cm. latis, basi late obtusis vel rotundatis, in acuminem gracilem obtusum ad 2 cm. longum gradatim angustatis, margine paullo recurvatis, subtus minute et dispersim glanduloso-strigillosis, basim versus 5-nerviis, costa et nervis secundariis (intimis cum costa ad 2 cm. concurrentibus) adscendentibus supra leviter impressis subtus elevatis, rete venularum supra plerumque immerso subtus prominulo; floribus axillaribus ut videtur solitariis, bracteis basalibus paucis obscuris; pedicellis gracillimis sub anthesi 20-30 mm. longis minute albido- puberulis basim versus minute bibracteolatis superne obscure glandulosis; calyce sub anthesi 4.5-5 mm. longo et apice diametro albido-puberulo, tubo cupuliformi 2.5-3 mm. longo basim versus obscure rubro-glanduloso, limbo erecto-patente sub- membranaceo circiter 2 mm. longo minute 5-dentato (dentibus apiculatis circiter 0.5 mm. longis), sinibus complanatis; disco annulari-pulvinato carnoso glabro; corolla tenuiter carnosa glabra anguste cylindrico-campanulata sub anthesi 20-22 mm. longa, basim versus circiter 4-mm. superne 7-8 mm. diametro, lobis 5 reflexis obtusis circiter 4 x 3 mm.; staminibus 10 longitudine corollam fere aequantibus, filamentis liberis membranaceis ligulatis circiter 13 mm. longis utrinque villoso- puberulis (pilis circiter 0.3 mm. longis), antheris 7-7.5 mm. longis, thecis levibus circiter 2.5 mm. longis basi rotundatis, tubulis 5-5.5 mm. longis per rimas ovales 1-1.5 mm. longas dehiscentibus; stylo filiformi corollam subaequante, stigmate minuto. Santiago-Zamora: Eastern slopes of the cordillera, valley of the Rio Negro, junction of Rios Pailas and Negro (on the trail to Mendez), 6,000-7,500 ft. elev., Aug. 20-24, 1945, Camp E-4924 (coll. F. Prieto) (TYPE US 1,989,113; dupl. NY) (shrub 3 m.; leaves deep green above, bright green and subnitid beneath), Eleutherostemon gracilipes is immediately distinguished from its congeners by its comparatively long pedicels and corollas and its greatly elongated filaments. In its solitary flowers and puberulent calyx it suggests the preceding new species (E. oliganthum), but in floral dimensions the two are very distinct. Disterigma alaternoides (H. B. K.) Nied. Bot. Jahrb. 11: 224. 1889. Cafiar: Northeast of Azogues, Camp E-1784, F. Prieto P-90, Azuay: Ridge between E] Pan and Guachapala, Camp E-5256, North of Paute, Camp E-2594, The eastern Cordillera, north of Sevilla de Oro, Camp E-4271, ‘‘Oriente’’ Border, Paramo del Castillo and surrounding forested areas (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), Camp E-720. Loja: Loma de Oro, near Saraguro, Camp E-558 (NY only), E-692. Nudo de Guagrauma, slopes of the Loma de Oro, Camp E-274 (NY only). Cerro Villanaco, west of Loja, Camp E-249, E-250, ‘‘Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, Camp E-74, E-90, Napo-Pastaza: Valley of the Rfo Pastaza and adjacent uplands, vicinity of E] Topo, Camp E-1684, E-1685, E-1690 (NY only), E-2410 (NY only). Santiago-Zamota: Eastern slope of the cordillera, valley of the Rfos Negro and Chupian za, Tambo Chontal to Tambo Consuelo, Camp E-1595 (NY only). ‘ 50 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol. 8, No. 1 The cited specimens were obtained at altitudes of 4,000 to 11,000 ft., the plant having been noted most often as a shrub, low and spreading in exposed places but up to 4 m. high where protected from winds; it is sometimes epiphytic, either de- pendent or with erect stiff branches; leaves deep green above, paler beneath, often nitid; corolla white to pink; mature fruit about 1 cm. in diameter, white or pinkish to reddish purple or wine-red, translucent. This excellent series of specimens, together with many others which have be- come available since my previous discussion of the species (Brittonia 1: 219, 220. 1933), show that D. alaternoides varies to such a degree that the variety parvifolium (Benth.) A. C. Smith no longer has any significance. This variety occurs throughout the range of the species (Venezuela to Bolivia) and is distin- guishable only on the basis of its smaller than typical leaves. It is now seen that leaves on different branches of the same plant are often quite diverse in size, their dimensions perhaps depending largely upon age or exposure. Although I originally noted the flowers as ‘‘glabrous,’”’ it should be noted that some of the specimens now available have both calyx and corolla sparsely pilose with minute whitish hairs, Disterigma leucanthum A. C. Smith, sp. nov. Frutex epiphyticus vel muscicola, ramulis gracilibus obtuse angulatis juventute minute puberulis demum glabratis; petiolis subteretibus rugulosis 1.5-2.5 mm. longis ut ramulis puberulis; laminis subcoriaceis in sicco fusco-viridibus ellipticis, 15-22 mm. longis, 8-15 mm. latis, basi rotundatis vel late obtusis, apice late obtusis, margine leviter recurvatis, juvenilibus parce puberulis mox glabratis, subtus parce et subpersistenter rubro-glanduloso-strigillosis, e basi obscure 5- nerviis, costa et nervis secundariis supra prominulis subtus obscuris vel leviter elevatis, rete venularum supra saepe prominulo subtus immerso; inflorescentia uniflora bracteis inconspicuis pluribus papyraceis imbricatis suborbiculari-oblongis parce ciliolatis (maximis circiter 1.5 mm. longis) basi circumdata; pedicellis bracteolis calyce et corolla extus dense et uniformiter puberulis (pilis pallidis patentibus circiter 0.1 mm. longis); pedicellis teretibus 1-2°mm. longis, bracteolis apicalibus late ovato-reniformibus circiter 1 mm. longis et 3 mm. latis ciliolato- marginatis imbricatis quam calycis tubo brevioribus; calyce sub anthesi circiter 5 mm. longo et apice diametro, tubo obscure angulato circiter 1.5 mm. longo, limbo erecto-patente profunde 4-lobato, lobis ovatis 2.5-3 mm. longis latisque interdum basi anguste imbricatis apice obtusis intus glabris; disco annulari-pulvinato glabro; corolla carnosa cylindrica sub anthesi circiter 9 mm. longa et 3.5 mm. diametro intus glabra, lobis 4 oblongis circiter 2 mm. longis obtusis; staminibus 8 quam corolla brevioribus, filamentis gracilibus liberis ligulatis 2-2.5 mm. longis superne intus parce villosis, antheris 4.5-5 mm. longis, thecis basi rotundatis parce hispi- dulis, tubulis thecas longitudine subae quantibus per rimas ovales 0.5-1 mm. longas dehiscentibus; stylo crasso tereti corollam subaequante, stigmate minuto. Santiago-Zamora: Cordillera Cutuct, ridge just south and west of Rio Itzintza, 5,000-5,900 ft. elev., Nov. 17-Dec. 5, 1944, Camp E-1344 (TYPE US 1,989,017; dupl. NY) (epiphyte, or at 5,900 ft. seen growing in a mound of sphagnum; corolla pure white). : | | The new species is a relative of D, alaternoides (H. B. K.) Nied., differing in its very short pedicellary bracteoles which do not effectively conceal the’calyx- tube (as they do in D. aluternoides, where they are usually 2-4 mm. long), its copiously and uniformly puberulent flowers (glabrous or very sparsely pilose in D. alaternoides), its large calyx-lobes (1-2 mm. long in D. alaternoides), and its longer corolla and anthers. Other species of this general alliance, D. popenoet Blake and D. ulei Sleumer, differ from the new species in many obvious characters, 1952] | PLANTS COLLECTED IN ECUADOR 51 the first having conspicuously nerved leaves, several-flowered inflorescences, a glabrous short-lobed calyx, and small anthers, the second having thick-carnose obovate leaves and very small flowers (calyx teeth minute; corolla about 3.5 mm. long, glabrous; anthers about 2 mm. long). Disterigma empetrifolium (H. B. K.) Drude in E.& P. Nat. Pfl. 4 (1): 52. 1889. Cahar: Uplands called ‘‘Huairacaja,’? 10 20 km. northeast of Azogues, Camp E-1777. Azuay: Along the Rio Matadero, west of Cuenca, Camp E-2021. Vicinity of the lake in the valley of the Rio Surucuchu (a branch of the Rio Matadero), 18-20 km. west of Cuenca, Camp E-4162, Paramo de Tinajillas and surrounding chaparral and forests, 30-50 km. south of Cuenca, Camp E-467. ‘‘Oriente’’ Border, Eastern Cordillera, between Ona andthe Rio Yacuambi, F. Prieto P-303. Loja: Cerro Villanaco, about 7 km. west of the city of Loja, Camp E-247. The species was often common where noted, at elevations of 8,000 to 11,200 ft., growing as a low shrub in grass, among rocks, on banks, or in bogs, often forming mats or dense clumps and propagating by runners; leaves dark green above, paler beneath, dull on both sides or shining above; corolla light rose or deep pink to crimson; filaments white, the anthers red-brown; mature fruit white, translucent, oblate-spherical, up to 1 cm. in diameter, insipid. This is the common small-leaved species of Disterigma, occurring along the Andes from Venezuela to Peru. Disterigma codonanthum Blake, Jour. Wash. Acad. 16: 363. 1926, Azuay: Cordillera de Alpachaca (headwaters of the Rio Jubones, between the Rios Giron and Leén), near the pan-American highway at about km. 79, Camp E-405. The eastern Cordillera, 4-6 km. north of the village of Sevilla de Oro, Camp E-4717A, E-4717B, ‘‘Oriente’’ Border, Paramo del Castillo and surrounding for- ested areas (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), Camp E-4869 (NY only); same locality, east of El Pan, Camp E-1632, The cited specimens were obtained at elevations of 9,000 to 11,350 ft.; they were collected at the edges of paramos and sometimes formed the dominant ground- cover on open slopes and paramo areas; in habit these plants often occur in dense mats in sphagnum meadows, propagating by long runners, sometimes almost buried in the sphagnum with only the tips exposed, or with the aerial parts stiffly erect and rarely as high as 0.5 m.; leaves shining, deep green above, paler beneath; bracts and calyx bright green; corolla green, red-tinged, brick-red, or rosy-pink (total range of color variation from green to red sometimes found in single plants); filaments white to bright pink, the anthers brown; fruit white, translucent. This species, apparently endemic to Ecuador, is less rare than indicated by the fact that I cited only two collections in 1933 (Brittonia 1: 228); I have since seen material from Carchi: (Penland & Summers 871, NY), Imbabura (Penland & Summers 818, NY), and Azuay (Steyermark 53436, A, Ch). Some of the corollas on the Camp specimens are larger than previously noted, being up to 10 mm. long. Nimber 4869 has leaf-blades notably larger than usual (up to 10 x 5 mm.), but its flowers are typical for the species. Number 1632, from the same general locality, . has one similarly large-leaved branch, from which arise lateral shoots with leaves normal for the species (1.5-3 mm. broad). These interesting specimens show what variable and undependable characters the shape and size of leaves are in this complex. Disterigma campii A. C. Smith, sp. nov. Frutex epiphyticus, ramulis elongatis gracilibus fusco-castaneis obtuse angu- latis vel subteretibus copiose hispidulis (pilis castaneis circiter 1 mm. longis subpersistentibus); foliis non confertis 4 vel 5 per centimetrum; petiolis teretibus |b % 2 V4 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 gracilibus circiter 1 mm. longis parce breviter hispidulis; laminis papyraceis in sicco supra olivaceis subtus fuscis oblongo-ellipticis, (10-)12-15 mm. longis, (3-)5-8 mm. latis, basi obtuse cuneatis, apice obtusis vel rotundatis, margine re- curvatis, utrinque glabris vel subtus raro obscure glandulosis secus costam, e basi obscure 5-nerviis, costa supra leviter impressa subtus elevata, nervis aliis immersis; floribus solitariis axillaribus subsessilibus praeter calycis lobos et filamentas glabris, bracteis basalibus paucis papyraceis suborbicularibus maximis haud 1.5 mm. diametro; pedicellis gracilibus circiter 0.5 mm. longis, bracteolis apicalibus papyraceis suborbicularibus 3-3.5 mm. longis et latis rotundatis calycis tubum amplectentibus; calyce sub anthesi circiter 5 mm. longo et apice diametro, tubo cupuliformi circiter 2 mm. longo, limbo erecto-patente papyraceo profunde 4-lobato, lobis oblongo-deltoideis 2.5-3 mm. longis 2-2.5 mm. latis apice incrassatis subacutis minute hispidulis; disco annulari-pulvinato; corolla tenuiter carnosa late campanulata etiam subrotata 5-6 mm. longa apice 7-8 mm. diametro profunde 4-lobata, lobis deltoideis 3.5-4 mm. longis et basi circiter 3 mm. latis valde re- flexis apice obtusis; staminibus 8 quam corolla brevioribus, filamentis ligulatis gracilibus liberis circiter 1.5 mm. longis superne parce hispidulis, antheris 3-3.5 mm. longis, thecis basi rotundatis, tubulis gracilibus quam thecis brevioribus per rimas ovales circiter 0.5 mm. longas dehiscentibus; stylo crasso tereti leviter exserto, stigmate minuto. Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, near junction of El Tigre and Pastaza, below Topo, 5,600 ft. elev., May 9, 1944, Camp E-1692 (TYPE US 1,989,039; dupl. NY) (epiphyte, blown from tree-top; terminal new growth in bud notably mucilaginous; flowers white; fruit white, translucent). The very distinct new species here described is characterized by its lax habit, elongate branchlets with a persistent hispidulous indument, well-spaced leaves of average size for the genus, solitary subsessile flowers, and especially by its broadly campanulate or even subrotate corollas. In size and shape of leaves it approximates D. humboldtii (K1.) Nied. (Guateniala to Venezuela and Colombia), but that species has shorter calyx-lobes, a cylindric corolla, comparatively long filaments, and anthers with proportionately longer tubules. In floral characters the new species seems closest to D. codonanthum Blake, and its leaves (although still considerably larger) are even suggestive of the large-leaved phase of that species, discussed above. However, D. campii also differs from D. codonanthum in its branchlet-indument, shorter pedicels, shorter and even more broadly cam- panulate corollas, and short filaments. Disterigma micranthum A. C. Smith, sp. nov. Frfitex nanus, ramulis primo brunneis obtuse angulatis laxe pilosis (pilis cir- citer 1 mm. longis) demum cinerascentibus teretibus glabratis, ramulis brevibus bracteis papyraceis imbricatis lanceolato-oblongis (maximis circiter 6 mm. longis) basi circumdatis; foliis non confertis 3 vel 4 per centimetrum; petiolis gracilibus subteretibus circiter 1 mm. longis parce puberulis; laminis papyraceis in sicco pallide olivaceis anguste ellipticis, 8-10 mm. longis, 3-4.5 mm. latis, basi et apice obtusis, margine subplanis, interdum apicem versus obscure puberulis et subtus parce glanduloso-strigillosis, e basi obscure 3- vel 5-nerviis, costa nervisque immersis vel.subtus leviter prominulis; floribus solitariis axillaribus subsessili- bus praeter filamentas glabris, bracteis basalibus paucis suborbicularibus maximis haud 1 mm. diametro; pedicellis gracilibus haud 1 mm. longis, bracteolis apicali- bus papyraceis suborbicularibus 2.5—3 mm. diametro scarioso-marginatis basi imbricatis calycis tubum amplectentibus; calyce sub anthesi 2.5-3 mm. longo et apice diametro, tubo obtuse angulato circiter 1.5 mm. longo, limbo erecto-patente 4-lobato, lobis deltoideis circiter 1 x 1.5 mm. subacutis; disco annulari-pulv inato; 1952] | PLANTS COLLECTED IN ECUADOR 53 corolla tenuiter carnosa urceolata sub anthesi 3.5-4 mm. longa et 2.5=3 mm. dia- metro basi et faucibus contracta, lobis 4 oblongis circiter 1 mm. longis subacutis; staminibus 8 quam corolla brevioribus, filamentis gracilibus liberis circiter 2 mm. longis superne villosis (pilis albidis circiter 0.4 mm. longis), antheris cir- citer 1.4mm. longis, thecis basi rotundatis, tubulis gracilibus thecas subaequantibus per rimas ovales circiter 0.5 mm. longas dehiscentibus; stylo tereti corollam subaequante, stigmate minuto. El Oro: In Moro-Moro region, about 21 miles west of Portovelo, 3,400-4,200 ft. elev., Oct. 7, 1944, Camp E-616 (TYPE NY) (in dense rain-forest, on banks; flowers white). : This new species is without close allies, being characterized by its very small flowers with urceolate corollas. ;In foliage it may most nearly suggest D. hum- boldtii (Kl1.) Nied. and D. campii (described above), but the size and shape of the corolla and the minute: anthers of D. micranthum make detailed comparisons superfluous. Disterigma sp. Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, low hills east of Puyo, 3,000 ft. elev., Camp E-1699, .E-1700 (both NY only) (epiphytes; fruit subtranslucent). The cited specimens represent another species of Disterigma with a lax habit, leaves comparatively spaced on the branchlets, and subsessile solitary flowers. They closely resemble the preceding new species (D. micranthum) in foliage, but the calyx-lobes are comparatively elongate (about 2 mm. long) and hispidulous- ciliolate on the margins; the leaves also are persistently ciliolate-margined. Although I feel certain that these two collections represent an undescribed species, in the absence of corollas I think it best to await more complete material. Disterigma acuminatum (H. B. K.) Nied. Bot. Jahrb. 11: 209. 1889. Pichincha: Western slope of the cordillera, along the road from Quito to Sto. Domingo de los Colorados, Camp E-1729 (NY only). Azuay: Paramo and sub- paramo area north and northwest of the Paramo del Castillo (6-8 km. north-northeast of Sevilla de Oro), Camp E-5173. Paramo del Castillo and surrounding forested areas (crest of the eastern cordiilera on the trail between Sevilla de Oro and Mendez), Camp E-4810. Loja: ‘‘Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, Camp E-75, E-96. Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, Sierra de los Leones, near Banos, Camp E-1696, The above collections come from elevations of 7,000-11,200 ft.; the plant is noted as an epiphyte, with stems pendant to 2 m., or as terrestrial, erect or spreading, up to 3 m. high, often irregularly branched; leaves pale green or deep green above and paler beneath; corolla greenish yellow or white tinged with pink; fruit white, translucent. Disterigma acuminatum is a well-marked and fairly abundant species, occur- ring along the Andes from Colombia to Peru. Some of the Ecuadorian specimens, such as Camp E-1729 and Sydow 608 (US), from Tungurahua, have leaves larger | than normal for the species. Disterigma pentandrum Blake, Jour. Wash. Acad. 16: 364. 1926. Chimborazo: Cafion of the Rio Chanchan, about 5 km. north of Huigra, Camp E-3313; same general locality, directly above the village of Huigra, Camp E-3480 (NY only). Cafiar: Valley of Rio de Cahar at ‘‘Selem,’’ between Galleturo and Canhar, F. Prieto CP-40. Azuay: Nudo de Portete, Pacific side of pass between headwaters of the Rios Tarqui and Giron, Camp E-2175. ‘‘Oriente’’ Border, east slope of Eastern Cordillera, between Ofia and the Rio Yacuambi, F. Prieto P-271. Y 54 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 Santiago-Zamora: Valley of the Rio Zamora, east of Loja,’ ridge across the river from the village of Zamora, Camp E-34 (NY only). The cited specimens were collected at elevations of 5,000 to 9,500 ft.; the plants were noted as epiphytes or as shrubs up to 4 m. high growing on cliffsides or hanging from rocks; leaves crisp- or inflated-succulent, 1-2 mm. thick, deep green above, paler beneath, usually shining on both sides or dullish beneath; bracts pale chestnut-brown; -calyx light pink to red; corolia crimson; fruit pale blue to deep lavender, the projecting calyx-lobes purplish or nigrescent-purple. This series of specimens is a valuable addition to the known material of the species, which in 1933 (Brittonia 1: 231) I had known only from the two original collections. Since that time Sleumer (Notizbl. Bot. Gart. Berlin 12: 123. 1934) has cited several additional Ecuadorian specimens and discussed variation within the species, which is considerable but no more than normal in Disterigma. Another specimen of this relationship is: Pichincha: Western slope of the cordillera, along the road from Quito to Sto. Domingo de los Colorados, about 6,000 ft. elev., Camp E-1738 (NY only) (plant hanging down to 0.3 m.; on cliff; flowers solitary in axils, deep rose; leaves subcarnose, red-tinged). This number is certainly allied to D. pentandrum, like which it has 5 stamens. However, its calyx-lobes are 1-1.5 mm. long, eciliate, and not thickened (those of the species being usually 2,5-4 mm. long, obviously glandular-ciliate, and distally thickened). I suspect that no. 1738 represents an undescribed species of this alliance, but the material is not adequate for description. Vaccinium floribundum H. B. K. Nov. Gen. & Sp. 3: 266. pl. 251. 1818. [Colombia: Narifio: Near Chiles, Camp E-341.] Carchi: Camp E-297, E-298, FE-313-E-317 incl., E-333. Pichincha: Camp E-1704, E-1714. Leon: Camp E-2351. Canar: Camp E-1796. Azuay: Camp E-384, E-389, E-463, E-1641, E-2063, E-2140, E-2264, E-5174, E-5243, s: n.; F. Prieto P-236A,. P-236B, P=269, P-=311, Loja; Camp E-95, E-272. | : | The specimens cited above, which give an excellent picture of the degree of variation in typical V. floribundum in Ecuador, are accompanied by notes too ex- tensive to be given here. In brief, the plant was found at elevations of 7,500 to 12,000 ft.; it was noted as a shrub up to 2 m. (or rarely as much as 3.5 m.) high, often low, spreading, or prostrate on paramos, in small! or extensive colonies, with underground burls up to 10 cm. in diameter; leaves deep green above, paler beneath, dull on both sides or subnitid above; hypanthium green.to purplish, or suffused with pink, sometimes subglaucous, the calyx-lobes sometimes red; corolla usually bright pink to red, sometimes white tinged with pink; fruit blue to black, glau- cous Of not. The varieties of V. floribundum accepted by Sleumer(Notizbl. Bot. Gart. Berlin 13: 129-132. 1936) seem of very questionable value, being based on size and shape of leaves. However, I have attempted to arrange the present collections in these varieties. Those cited above fall into the typical variety as interpreted by Sleumer. It may be noted that nos. 389, 2264, 5243, and s. n. represent the form which Blake has described as V. dasygynum (which Sleumer reduces outright to typical V. floribundum), of which the calyx is pubescent. : Vaccinium floribundum var. ‘marginatum (Dun.) Sleumer, Notizbl. Bot. Gart. Berlin 1327 13). L956 Cafar: Northeast of Azogues, F. Prieto P-153. Azuay: Paramo de Tinajillas and surrounding chaparral and forests, south of Cuenca, Camp E-485. Cordillera de Alpachaca, Camp E-284 (NY only). The specimens are from elevations of 9,800 to 11,000 ft., and the plant is noted as a spreading or arching shrub to 0.2 m., with a pink corolla and black fruit. This ae 1952] PLANTS COLLECTED IN ECUADOR 55 variety, characterized by having its leaf-blades ovate and broadest toward the base, hardly seems worthy of separation from the typical variety; however, our specimens agree well with those from Ecuador cited by Sleumer. Vaccinium floribundum var. ramosissimum (Dun.) Sleumer, Notizbl. Bot. Gart. Berlin 13: 131. 1936. Carchi: Paramo del Angel, about 24 km. southwest of Tulcan, 11,500 ft. elev., Camp E-286, E-290 (NY only). Loja: Cerro Villanaco, about 7 km. west of Loja, 8,000-9,500 ft. elev., Camp E-210 (NY only). The specimens are from low shrubs, up to 0.3 m. high, the older ones arising from a characteristic burl; corolla pink; fruit dark blue, subglaucous. The variety ramosissimum, characterized by its very small leaves, is perhaps somewhat stronger than var. marginatum, but still it appears to be merely an extreme form of the species from exposed locations. Vaccinium crenatum (Don) Sleumer, Notizbl. Bot. Gart. Berlin 12: 291. 1935. Azuay: Paramo de Tinajillas and surrounding chaparral and forests, 30-50 km. south of Cuenca, Camp E-2284. ‘*‘Oriente’’ Border, Paramo del Castillo and sur- rounding forested areas (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), Camp E-1643, E-4819. Ridge between El Pan and Guachapala, Camp E-5245. Loja: Cerro Villanaco, about 7 km. west of Loja, Camp E-165, E-260 (NY only). Nudo de Cajanuma, south of Loja, Camp E-562 (NY only). Santiago- Zamora: East of El] Pan at about Azuay line, near ‘‘Laguna,’’ F. Prieto P-62. The cited specimens were obtained at elevations of 7,500 to 11,500 ft.; the plants are usually prostrate and trailing (rarely erect to 1 m. high) on banks, open slopes, and bare eroding areas, often rooting along the stem; leaves often reddish when young, later deep green and shining above, with nerves reddish beneath; corolla pink to bright red or crimson; fruit nigrescent-blue or purple-black, shining. Venezuela to Peru; our material agrees well with the Ecuadorian specimens cited by Sleumer in Notizbl. Bot. Gart. Berlin 13: 133 (1936). Semiramisia speciosa (Benth.) Kl. Linnaea 24: 25. 1851. Loja: ‘*Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, ca. 10,000 ft. elev., Camp E-107 (scrambling shrub, growing to 5 m. high, with stems up to 2.5-3 cm. diam.; corolla cylindric, not apically fluted, bright red, the base near calyx yellowish green; calyx green, conspicuously fluted). The cited collection comes from near the type locality; I have recently (Contr. U. S. Nat. Herb. 29: 359. 1950) cited other Ecuadorian collections of the species. Semiramisia weberbaueri Hoer. Bot. Jahrb. 42: 310. 1909. Azuay: ‘‘Oriente’’ Border, Paramo del Castillo and surrounding forested area (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), 9,000-11,000 ft. elev., Camp E-735 (shrub to 2 m., on ground; corolla basally green, apically bright red). Santiago-Zamora: Easter slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), 6,500-7,500 ft. elev., Camp E-756 (shrub 4 m.; corolla basally green, apically Crimson). The differences between this species and the preceding are perhaps not very Significant; S. weberbaueri has comparatively narrow leaves which are acute to obtuse at base. The difference in the calyx-tube-whether smooth or angled=which I utilized in Contr. U. S. Nat. Herb. 28: 348-349 (1932), appears on the basis of more ample material to be of little consequence, the angles tending to disappear with maturity. The type of S. weberbaueri was from the Department of Amazonas, Peru, but other collections from Ecuador have been mentioned by me (op. cit. 350). \. 56 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN (Vol. 8, No. 1 Semiramisia hypogaea A. C. Smith, sp. nov. Frutex repens caespite fere saepe occultus praeter flores ubique glaber, ramu- lis gracilibus teretibus brunneis inferne radicantibus, internodiis interdum pauci- bracteatis, bracteis inconspicuis papyraceis lanceolatis circiter 2 mm. longis; petiolis subteretibus valde rugulosis 2-5 mm. longis crassis (1.5-2.5 mm. diametro); laminis crasso-coriaceis in sicco fusco-olivaceis late ovatis, 3.5-7 cm. longis, 2.5-4.6 cm. latis, basi rotundatis vel late cuneatis, apice subacutis raro (minori- bus) subrotundatis, margine incrassatis et paullo recurvatis, maturis inconspicue immerso-glandulosis vel sparsissime castaneo-glanduloso-strigillosis (juvenilibus pilis rubris glandulosis 0.2-0.3 mm. longis utrinque strigillosis), costa utrinque leviter elevata, nervis primariis basalibus plerumque utrinsecus 2 curvato- adscendentibus haud elevatis, venulis immersis; floribus paucis in axillis soli- tariis vel infra folia orientibus ubique (i. e. pedicello, calyce, et corolla) pilis albidis circiter 0.3 mm. longis patentibus indutis, bracteis minutis; pedicellis leviter curvatis sub anthesi 7-8 mm. longis basim versus minute bibracteolatis superne paullo incrassatis et cum calyce continuis; calyce 4.5-5.5 mm. longo | apice 4—5 mm. diametro, tubo cupuliformi circiter 3 mm. longo, limbo suberecto 1.5-2.5 mm. longo, lobis 5 deltoideis circiter 1.5 mm. longis et 1.5=2 mm. latis acutis, sinibus subacutis; disco annulari-pulvinato glabro; corolla siccitate fragili cylindrica sub anthesi 20-23 mm. longa 4-5 mm. diametro, lobis 5 oblongo-deltoideis 3.5-4 mm. longis 2=2.5 mm. latis obtusis; staminibus 10 corollam subaequantibus, filamentis ligulatis liberis glabris circiter 3 mm. longis, thecis levibus 5-6 mm. longis basi obtusis in tubulos gracillimos 15-18 mm. longos poris subterminalibus apertos terminantibus; stylo filiformi corollam subaequante, stigmate minuto. Loja: ‘‘Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, about 10,000 ft. elev., July 2, 1944, Camp E-88 (TYPE NY; dupl. US) (in pass, the plants spreading by runners, almost hidden in the short grass; young leaves brilliant red, green with age, and dull on both surfaces; flowers noted on old wood or in the axils of last year’s leaves, apparently solitary; corolla deep red, hidden in the grass or sometimes below the surface of the turf; when flowers are produced below the surface of the soil the pedicels elongate and bring them to the surface; anthers brilliant yellow; this plant would be easily missed were it not for the bright red of the young leaves). The remarkable habit described in Dr. Camp’s field notes, which indicate that the flowers are sometimes subterranean, has not otherwise been noted in this group. Perhaps, however, the related S. fragilis A. C. Smith [including Ceratostema longepedicellatum Sleumer; see Bull. Torrey Club 63: 312 (1936) for reduction], also ftom Ecuador, may have a similar habit. These two species form a very dis- tinct group without close allies in Semiramisia. From S. fragilis the new species differs in matters of degree which seem worthy of specific recognition; the leaves are considerably larger and predominantly acute at apex, the pedicels are shorter (but perhaps elongating as implied in the field note), the calyx is slightly larger, and the thecae of the anthers longer. Comparable dimensions in S. fragilis are: petioles 1.5-2 mm. long; leaf-blades 1.5-3 by 1.1-2.2 cm., rounded or broadly obtuse at apex; pedicels 10-20 mm. long at anthesis, up to 40 mm. in fruit; Selva 3-3.5 mm. long; thecae 2—3 mm. long. Ceratostema Juss. Gen. Pl, 163. 1789. Englerodoxa Hoer. Bot. Jahrb. 42: 310. 1909. Periclesia A. C. Smith, Contr. U. S. Nat. Herb. 28: 357. 1932. In my treatment of 1932 I recognized the genus Englerodoxa as composed of three species; shortly afterward the identity of this concept with Ceratostema 1952] PLANTS COLLECTED IN ECUADOR 57 (sensu vero, non sensu A. C. Smith, Contr. U. S. Nat. Herb. 28: 335-348. 1932) was pointed out by Sleumer (Notizbl. Bot. Gart. Berlin 12: 278-282, 1935) and my- self (Bull. Torrey Club 63: 307-309. 1936). Periclesia was established by the writer on the basis of a single species. At that time the distinction between the two genera seemed adequate, Periclesia having 4-merous flowers with extremely large calyx-lobes and connate filaments. Now, however, several additional species have been referred to Ceratostema and four additional ones to Periclesia. These species serve effectively to break down the differences originally believed to separate the two genera, and the exten- sive material of this complex assembled by Dr. Camp and his assistants further indicates that the two concepts are no longer useful. With the addition of three novelties herewith described, 16 species of Ceratostema (sensu vero) may now be recognized. In order to facilitate identification in this difficult genus I give below a key to the known species. Ceratostema may be circumscribed as having the following fundamental char- acters: calyx articulate with pedicel (the articulation rarely obscure or even lacking, in C. loranthiflorum); corolla large, often ventricose near base, deeply lobed; filaments free or connate, glabrous or pilose (but never with massed retrorse hairs); stamens with strongly granular thecae and very slender stiff tubules which dehisce by short oblique subterminal pores. It is geographically limited to the Andean area extending from southern Colombia through Ecuador and possibly into northern Peru, the precise locality of some collections being questionable. In the Ecuadorian Andes Ceratostema seems to be one of the most frequent and certainly one of the most striking vacciniaceous constituents of the flora. In reconsidering the genus Ceratostema mention should be made of C. speciosum André (Illustr. Hort. 17: 52. pl. 9. 1870; A. C. Smith, Contr. U. S. Nat. Herb. 28: 345. 1932), which I referred to the genus Plutarchia in 1936 (Bull. Torrey Club 63: 312). The type of the species was obtained near Loja (south of the usual range of Plutarchia), and from the inadequate original description and plate it seems possible that the species actually does represent Ceratostema in the sense of the present treatment. The important character of the anther-dehiscence cannot be ascertained from the original publication; lacking this, I am still unable to place the species with certainty. If it does fall into Ceratostema it appears distinct from any of the species in my key below. KEY TO THE SPECIES OF CERATOSTEMA Calyx-limb very conspicuous, the lobes elongate-deltoid, at anthesis at least 10 mm. long. Leaf-blades deeply cordate at base; calyx-lobes membranaceous, conspic- uously reticulate-nerved. Rachis, pedicels, and calyx pilose with whitish eglandular hairs} bracts and bracteoles small; calyx-tube 10-costate; leaf-blades soft-pilose beneath. , C. peruvianum Gmel. Rachis, pedicels, and calyx copiously pilose with weak gland-tipped hairs; bracts and bracteoles papyraceous, 5—17 mm. long; calyx- tube terete; leaf-blades essentially glabrous. C. pensile A. C. Smith, comb. nov. Leaf-blades attenuate to obtuse at base. : Calyx-tube obconical, smooth, the lobes membranaceous, conspicuously reticulate-nerved; anthers with thecae 5-6 mm. long; flowers 4- merous; leaf-blades 4—6 x1.2-1.8 cm. C. flexuosum (A. C. Smith) Macbr. Calyx-tube costate or winged, the lobes chartaceous to coriaceous; an- thers with thecae 8-18 mm. long; flowers 5-merous (calyx- and corolla-lobes sometimes partially fused). ‘ 58 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN (Vol. 8, No. 1 Leaf-blades lanceolate-oblong, 5-7 x 1—2 cm., obscurely pinnatinerved, soft-pilose beneath; pedicels, calyx-tube, and corolla densely pale-pilose; calyx-tube winged. C. lanceolatum Benth. Leaf-blades elliptic or ovate-elliptic (rarely lanceolate), 6-12 x 2-8 cm., essentially glabrous, the secondary nerves ascending from or nearly from base; flowers glabrescent at or soon after anthesis. Calyx 30—45 mm. long, the tube conspicuously 5-winged. C. reginaldii A. C. Smith, comb. nov. Calyx 20-28 cm. long, the tube regularly 10-costate. C. alberti-smithii (Sleumer) Sleumer. Calyx-limb comparatively inconspicuous, the lobes deltoid, ovate, or merely apiculate, not more than 7 mm. long at anthesis. Sinuses of calyx-limb usually acute or obtuse, the lobes obvious, 1-7 mm. long. Leaf-blades conspicuously cordate and amplexicaul at base. Calyx-tube 10-costate; corolla 25-30 mm. long; leaf-blades up to 7.5 x 4.5 cm.; branchlets and young leaves soft-pilose. C. amplexicaule A. C. Smith. Calyx-tube terete; corolla 33-35 mm. long; leaf-blades up to 13.5 x 7 cm.; branchlets and leaves glabrous. C. silvicola A. C. Smith. Leaf-blades acute to rounded or faintly cordate at base, not amplexicaul. Corolla 15=20 mm. long; calyx continuous with pedicel (or the articu- lation very obscure); leaf-blades usually lanceolate-elliptic, rarely more than 3.5 cm. broad. C. loranthiflorum Benth. Corolla at least 35 mm. long; calyx obviously articulate with pedicel. Leaf-blades 6-12 cm. long, the principal nerves sharply ascending from or nearly from base. Calyx-tube longitudinally furrowed, the lobes about 6 mm. long; corolla about 6 mm. in diameter near base; filaments about 3 mm. long. C. calycinum (A. C. Smith) Sleumer. Calyx-tube smooth, the lobes 2=3.5 mm. long; corolla strongly ventricose, 8-17 mm. in diameter near base; filaments about 5 mm. long. C. ventricosum A. C. Smith, sp. nov. Leaf-blades 2-6 cm. long, pinnatinerved. | Leaf-blades serrate or at least crenulate at the usually strongly recurved margin; calyx-tube narrowly winged, the limb scarce- ly 2 mm. long including the lobes. C. alatum (Hoer.) Sleumer. Leaf-blades entire at margin; calyx-lobes 5-7 mm. long. Calyx-tube and corolla essentially terete, the calyx-lobes with marginal glands; corolla essentially glabrous; filaments firmly connate, the tubules 3—4 times as long as the thecae. C. nubigenum A. C. Smith, comb. nov. Calyx-tube and corolla 5-angled, the calyx-lobes eglandular; corolla villose; filaments free, the tubules 2-3 times as long as the thecae. C. campii A. C. Smith, sp. nov. Sinuses of calyx-limb flattened, the teeth minute, apiculate, scarcely 1 rim. long. Calyx-tube 5-winged; corolla about 30 mm. long, essentially hypocrateri- form, the limb flaring, with lobes 7-8 mm. broad at base; leaf-blades Ppinnatinerved, with 4-6 secondaries per side. C. charianthum A. C, Smith. Calyx-tube not winged; corolla 45-53 mm. long, subcylindric, the lobes about 3 mm. broad at base; leaf-blades 5- or 7-nerved from or nearly from base. : C. prietoi A. C. Smith, sp. nov. Ceratostema pensile (A. C. Smith) A. C. Smith, comb. nov. Periclesia pensilis A. C. Smith, Contr. U. S. Nat. Herb. 29: 364. 1950. As mentioned in the above discussion, the connate filaments hardly serve to keep Periclesia apart from Ceratostema. In its more fundamental characteristics the present species seems closest to C. peruvianum, differing in its glandular-pilose inflorescence with large bracts and bracteoles, as well as in its essentially glabrous leaves. Thus far C. pensile is known only from the type collection, Steyermark 53798. 1952] PLANTS COLLECTED IN ECUADOR 59 Ceratostema flexuosum (A. C. Smith) Macbr. Univ. Wyom. Publ. 11: 42. 1944. Periclesia flexuosa A. C. Smith, Contr. U. S. Nat. Herb. 28: 357. pl. 7. 1932. The species upon which the genus Periclesia was founded is also seen to be inseparable from Ceratostema, in view of subsequently described species of the complex. The 4-merous flowers and connate filaments can now hardly be considered of generic significance. The species remains known only from the type, Lobé 79, of which the precise locality. is uncertain. Ceratostema reginaldii (Sleumer) A. C. Smith, comb. nov. Periclesia reginaldii Sleumer, Bot. Jahrb. 71: 400. 1941. Ceratostema macranthum A. C. Smith, Contr. U. S. Nat. Herb. 29: 361. 1950. Azuay: The eastern Cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000-9,000 ft. elev., Camp E-4342, E-4619. **Oriente’’ Border, Paramo del Castillo and surrounding forested areas (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), 9,000-11,000 ft. elev., Camp E-722, E-4851. Loja: **Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, ca. 10,000 ft. elev., Camp E-108. Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), Tambo Chontal to Tambo Consuelo, 5,700-8,000 ft. elev., Camp E-1579. Field notes indicate that the species is diverse in habit, being a climbing vine (sometimes epiphytic), an erect or spreading shrub, or a tree up to 6 m. high; - the leaves are deep green and shining above, paler and dull beneath; the inflo- rescence parts are deep crimson to red, the corolla being somewhat paler within and sometimes with salmon-pink lobes; the fruit is green or pale yellowish and is noted as being eaten by birds. In recently proposing the new species C. macranthum I failed to consider the description of Periclesia reginaldii, of which type material is not available to me. However, a careful perusal of Sleumer’s description indicates that the two species are identical. Sleumer referred his plant to Periclesia because of the connate filaments; these are, even in the type of C. macranthum and especially in some of the Camp specimens cited above, firmly coherent in young flowers, becoming at length essentially free. The type of C. reginaldii is R. Espinosa.785, collected near Loja. The six collections cited above form a remarkable accretion to the known material of the species, which, as might be expected, proves more variable than indicated in the two previous descriptions. The following emendations should be noted. All of the Camp collections prove to have a much more fugacious .indument than de- scribed for C. macranthum, or, indeed, they are essentially glabrous throughout. | The leaves and inflorescence parts (including calyx and corolla) are in the present series glabrous at anthesis, or the calyx-limb may be puberulent only within. ‘As. to leaf-shape, nos. 108, 4342, and 4851 are very similar to Steyermark 54311, the type of C. macranthum. Numbers 722 and 4619 have leaf-blades more or less ovate and rounded at base, attaining dimensions of 12 x 8 cm. The inflorescence, in the Camp series, sometimes has as many as 15 flowers (although fewer usually develop) and a rachis up to 22 cm. long. The pedicel-length is very variable, sometimes © 15-75 mm. on the same specimen (e. g. no. 722). Slight extensions of floral di- mensions may be noted as follows: Calyx sometimes as short as 30 mm. (25 mm. on no. 1579) at anthesis, with lobes rarely as short as 10 mm.; corolla up to 55 mm. in length, with lobes up to 25 mm., these sometimes partially fused into 3 or 4 instead of 5; anthers up to 45 mm. (or perhaps more) in length, with thecae up to 20 mm. The most extreme specimen of those cited is no. 1579, which has the leaf-blades nearly lanceolate, 6-7 x 2-3 cm. (i. e. much narrower than usual), and the calyx in young fruit comparatively short. | — ‘ 60 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 Ceratostema loranthiflorum Benth. Pl. Hartw. 142. 1844. Loja: Cerro Villanaco (ca. 7 km west of the city of Loja), 8,000-9,500 ft. elev., Camp E-163, E-164, E-174 (NY only), E-175 (NY only), E-181, E-197, E-674. This species has thus far been collected only in the Province of Loja, and even there it has been known from only half a dozen collections. Dr. Camp’s new material, therefore, is very welcome as permitting a better understanding of varia- tion within the species. The plants are noted as shrubs from 0.3 to 2 m. high, arising from large subterranean burls. The leaves and flowers are very variable in width; e. g. nos. 175 and 181 have leaves up to 3.5 cm. broad, whereas in other available specimens they hardly exceed 2 cm. in breadth. Dr. Camp notes that the short broad type of leaf is correlated with a broad corolla, while relatively narrow leaves are associated with narrow corollas. No appreciable differences in corolla- length were noted in the various types. Frequently each calyx-lobe bears an in- conspicuous callose gland dorsally near its base. Ceratostema ventricosum A. C. Smith, sp. nov. Frutex epiphyticus interdum scandens, :ramulis gracilibus juventute albido- puberulis mox glabratis cinerascentibus; stipulis intrapetiolaribus inconspicuis oblongis circiter 2 mm. longis acuminatis; petiolis subteretibus rugulosis 2-6 mm. longis ut ramulis primo puberulis; laminis glabris in sicco fusco-viridibus lanceolato- vel ovato-ellipticis, 6-12 cm. longis, 2-5.3 cm. latis, basi obtusis vel rotundatis, in acuminem circiter 1 cm. longum gradatim angustatis, margine haud recurvatis, 5- vel obscure 7-nerviis, nervis intimis ad 1 cm. concurrentibus et costa supra leviter elevatis vel impressis subtus subprominentibus, nervis extimis inconspicuis, rete venularum utrinque prominulo vel supra immerso; inflorescentiis apices ramu- lorum versus axillaribus racemosis 2-8-floris breviter pedunculatis, rhachi tereti rugulosa 0.8-3 cm. longa obscure albido-puberula, bracteis sub floribus deltoideis 1.5-2.5 mm. longis acutis extus puberulis margine glandulas rigidas circiter 0.2 mm. longas interdum gerentibus; pedicellis teretibus ut rhachi puberulis sub anthesi 13-27 mm. longis articulationem versus incrassatis, paullo infra medium bracteolis 2 bracteis similibus 1.5-2 mm. longis ornatis; calyce sub anthesi 5-7 mm. longo apice 7-9 mm. diametro extus parce puberulo, tubo cylindrico 2-3 mm. longo, limbo subpatente 5-lobato, lobis oblongo-ovatis 2-3.5 mm. longis circiter 3 mm. latis apice apiculatis margine paullo incrassatis saepe glandulam callosam nigrescentem dorso gerentibus, sinibus acutis vel obtusis; disco carnoso annulari- pulvinato glabro; corolla glabra carnosa urceolato-cylindrica sub anthesi 35-47 mm. longa, basim versus conspicue ventricosa et 8-17 mm. diametro, superme angustata, profunde 5-lobata, lobis subulato-lanceolatis circiter 20 mm. longis et basi 3”mm. latis; staminibus 10 corollam fere aequantibus, filamentis submem- branaceis liberis vel basim versus subcohaerentibus circiter 5 mm. longis intus parce pilosis, antheris 30-35 mm. longis, thecis 12-19 mm. longis basi obtusis et leviter incurvatis, tubulis gracillimis longitudine thecas subaequantibus per poros obliquos circiter 0.5 mm. longos dehiscentibus; stylo filiformi corollam subaequante, Stigmate minute. Santiago-Zamora: Eastem slope of the cordillera, near junction of Rios Pailas and Negro, 6,000-7,500 ft. elev., Aug. 20-24, 1945, Camp E-4912 (coll. F. Prieto) (TYPE US 1,989,112; dupl. NY) (basal tuber epiphytic, the branches to 1 m. long; leaves deep green, subnitid above, dull beneath; pedicels bright green, clavate; hypanthium pale green to dull reddish, the articulation well marked, usually also with a dark line; calyx-lobes usually reddish, the pitted glands nigrescent; base of corolla deep crimson, the lobes nigrescent, spreading). Easter slope and crest of main Cordillera Cutuct, 5,200 ft. elev., Jorgensen CuJ]-39 (NY only) (epiphytic vine; flowers red). Cordillera Cutucu, ridge just south and west of Rio Itzintza, 1952] : PLANTS COLLECTED IN ECUADOR 61 4,500-5,500 ft. elev., Camp E-1333 (NY only) (high-growing epiphyte; corollas found on ground, crimson, the lobes purple-black). Of the cited specimens, no. 1333 consists only of fallen corollas, but it may confidently be referred to this species. The other two specimens show some foliar variability, the leaf-blades of no. CuJ-39 tending to be more ovate and broader than those of the type, with rounded rather than obtuse bases. In inflorescence characters the two specimens are essentially identical, except that No. Cu]-39 has braets and bracteoles without the marginal glands which occur on the type; both specimens usually have a dorsal gland on each calyx-lobe of a very char- acteristic type. Ceratostema ventricosum, although a very distinct species, has characteristics suggestive of several of its allies. Its leaves and inflorescences resemble those of C. prietoi, described below, but the calyx-limb is strikingly different, the corolla is more obviously ventricose, and the anther-proportions are different. In its calyx the new species resembles C. loranthiflorum, but that species has the calyx-articulation obscure or lacking, the corolla and stamens much shorter, and the leaves very different. From C. calycinum, which it resembles somewhat in foliage, C. ventricosum differs in its smooth calyx-tube and much shorter and differently shaped lobes, its more strongly ventricose corolla, and its longer filaments. Ceratostema alatum (Hoer.) Sleumer, Notizbl. Bot. Gart. Berlin 12: 281. 1935. Azuay: Paramo and sub-paramo area north and northwest of the Paramo del Castillo (ca. 6-8 km. n.-ne. of Sevilla de Oro), 10,000-11,200 ft. elev., Camp E-5153. ‘‘Oriente’’ Border, Eastern Cordillera, between Ofia and the Rio Yacuambi, on crest,-10,000-11,200 ft. elev., F. Prieto P-295. ‘‘Oriente’’ Border, Paramo del Castillo and surrounding forested areas (crest of the eastem cordillera on the trail between Sevilla de Oro and Mendez), 9,000-11,000 ft. elev., Camp E-700, E-746; same locality, 11,000—-11,300 ft. elev., Camp E-4867; same locality, east of El Pan, 11,000-11,350 ft. elev., Camp E-1626. Field notes indicate the plant as a low shrub, up to 3 m. high in sheltered places; leaves dark green and shining above, pale and dull beneath; pedicels and hypanthium deep crimson or red; corolla deep crimson at base, shading to purple or black at tips of lobes; filaments pink; anthers brown; ripe fruit oblate-spheroid, up to 1.5 cm. long and 2 cm. in diameter, dull reddish, nitid, slightly sweet. The excellent series of specimens cited above nearly doubles. the number of collections known for this species, but the variation does not notably extend the limits of my earlier description (as Englerodoxa alata Hoer. in Contr. U.S. Nat. Herb. 28: 350. 1932). The leaves are sometimes up to 3 cm. broad and the corolla may be as short as 35 mm. at anthesis, although more often it exceeds 40 mm. in length. Since my 1932 treatment the following collections have been noted: Prov- ince of Pichincha, Acosta-Solis 8304 (Ch,,US); Tungurahua, Penland & Summers 310 (NY); Santiago-Zamora, Steyermark 54333 (Ch). Ceratostema nubigenum (A. C. Smith) A. C. Smith, comb. nov. Periclesia nubigena A. C. Smith, Contr. U. S. Nat. Herb. 29: 366. 1950. This is another species which, because of its connate filaments, I originally referred to Periclesia. Although, as originally noted, it has certain features sug- gestive of Periclesia pensilis (e. g. Ceratostema p.), it is a very distinct species closely related only to the following new entity. Ceratostema campii A. C. Smith, sp. nov. | Frutex interdum epiphyticus et scandens, ramulis teretibus primo pilis albidis 0.3-0.5 mm. longis villoso-puberulis mox glabratis; stipulis intrapetiolaribus e 1 62 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 basi incrassato subulatis 2-4 mm. longis; petiolis rugulosis crassis 2-5 mm. longis mox glabratis; laminis siccitate fuscis subcoriaceis lanceolato- vel oblongo- ovatis, 2.5-4.5 cm. longis, 1-3 cm. latis, basi rotundatis vel leviter cordatis, apice obtusis vel subacutis, margine valde recurvatis, utrinque primo pallide pub- erulis etiam parce glanduloso-strigillosis mox glabrescentibus, pinnatinerviis, costa supra plana vel leviter impressa subtus elevata, nervis lateralibus princi- palibus utrinsecus 2 vel 3 basim versus orientibus curvatis utrinque elevatis vel inconspicuis, rete venularum subimmerso; inflorescentiis axillaribus breviter racemosis 2-4-floris, rhachi angulata 4-6 mm. longa pilis 0.2-0.3 mm. longis albido-puberula, bracteis sub floribus deltoideis 1.5-3 mm. longis acutis extus puberulis; pedicellis sub anthesi 10-16 mm. longis parce villosis basim versus bibracteolatis, bracteolis bracteis similibus, articulatione manifesto; calyce sub anthesi 10-12 mm. longo et apice diametro, tubo circiter 5 mm. longo albido-villoso valde 5-angulato, limbo papyraceo suberecto ad basim 5-lobato extus parce villoso intus glabro utrinque inconspicue luteo-glanduloso, lobis ovato-deltoideis acumi- natis 5-7 mm. longis 4-6 mm. latis, sinibus acutis; disco carnoso cupuliformi glabro; corolla carnosa 5S-angulata, sub anthesi 37-45 mm. longa et inferne circiter 10 mm. diametro superne angustata, ut calyce albido-villosa, lobis 5 elongato- deltoideis subacutis; staminibus 10 corollam fere aequantibus, filamentis inter se liberis submembranaceis ligulatis 3-4 mm. longis glabris, connectivo superne gradatim angustato, antheris 30-42 mm. longis, thecis 11-12 mm. longis basi sub- acutis, tubulis gracillimis quam thecis longioribus per poros obliquos circiter 0.5 mm. longos dehiscentibus; stylo filiformi longitudine corollam subaequante, stig- mate truncato. Loja: ‘‘Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, ca. 10,000 ft. elev., July 2, 1944, Camp E-106 (TYPE US 1,988,934; dupl. NY) (stiff shrub 3 m. high; leaves dull on both surfaces; calyx conspicuously fluted; corolla carnose, deeply grooved, red, apically tinged with yellow). Nudo de Guagrauma, ca. 12 km. south of Zaraguro, 9,500-10,500 ft. elev., Camp E-134 (NY only) (climbing €piphyte in sotobosque; flowers pendulous; calyx angled; corolla angled, pale crimson, the tip greenish yellow). The two cited specimens agree very closely in fundamental details, but no. 134 has somewhat the narrower leaves and the larger flowers. The floral dimensions are probably a matter of age; in the description the larger dimensions are probably to be taken as representative of the floral measurements at anthesis. The new species superficially resembles C. nubigenum, differing, as noted in my key, in its villose flowers, its angled calyx-tube and corolla, the absence of marginal calycine glands, its free filaments, and the differently proportioned anther- thecae and -tubules. Ceratostema charianthum A. C. Smith, Contr. U. S. Nat. Herb. 29: 360. 1950. Santiago-Zamora: Cordillera Cutuct, ridge just south and west of Rio Itzintza, 4,500-5,500 ft. elev., Camp E-1334 (NY only). Eastern slope and crest of main Cordillera Cutuct, 5,600 ft. elev., Jorgensen CuJ-44 (NY only). Eastern slopes of the cordillera, near junction of Rfos Pailas and Negro, 6,000-7,500 ft. elev., Camp E-4932 (coll. F. Prieto). F ield notes describe the species as a weak epiphyte or a vine growing on mossy banks or climbing trees in mossy forest, with adventitious roots which become en- larged and tuberous near the stem; leaves deep green above, paler beneath, ee on both sides; calyx pinkish; corolla bright rosy pink to pale crimson. The three cited specimens are welcome additions to the material of the species, otherwise known only from the type, obtained in the same general region. Slight 1952) PLANTS COLLECTED IN ECUADOR 63 amplifications of the original description may be noted: leaf-blades up to 13.5 cm. long, sometimes as narrow as 1.5 cm.; rachis of inflorescence sometimes insig- nificant, only 2 mm. long; pedicels varying from 5 to 13 mm. in length; calyx some- times shorter than previously described, only 7 mm. long, the wings more obvious (about 1 mm. broad); corolla-lobes as much as 11 mm. in length; filaments connate toward base rather than completely free. Ceratostema prietoi A. C. Smith, sp. nov. Frutex epiphyticus ubique praeter filamentas glaber, ramulis gracilibus teretibus fuscis cinerascentibus; petiolis subteretibus rugulosis 3-5 mm. longis; laminis in sicco subcoriaceis vel papyraceis fusco-viridibus elliptico-lanceolatis, 10-17 cm. longis, 3-6 cm. latis, basi anguste rotundatis, in acuminem gracilem 1-2 cm. longum gradatim angustatis, margine integris et anguste recurvatis, 5(vel 7-)- nerviis, nervis e basi orientibus vel interioribus interdum ad 1-2 cm. concurrenti- bus, costa nervisque principalibus supra leviter impressis subtus elevatis, rete venularum utrinque subimmerso vel prominulo; inflorescentiis e ramulis infra folia orientibus racemosis 3-12-floris, rhachi subtereti 0.5-3 cm. longa breviter pedun- culata, bracteis sub floribus deltoideis circiter 1 mm. longis mox caducis; pedicellis teretibus 13-20 mm. longis superne incrassatis, basim versus bracteolis 2 deltoideo- subulatis circiter 0.7 mm. longis caducis ornatis, articulatione conspicuo; calyce sub anthesi 7-9 mm. longo et apice diametro, tubo cupuliformi circiter 5 mm. longo ruguloso haud angulato, limbo subererecto papyraceo quam tubo breviore 5-dentato, dentibus apiculatis 0.5-1 mm. longis, sinibus complanatis; disco carnoso annulari- pulvinato; corolla carnosa cylindrica 45-53 mm. longa, basim versus 7-9 mm. diametro, superne angustata, profunde 5-lobata, lobis deltoideo-lanceolatis circi- ter 15 mm. longis et basi 3 mm. latis; staminibus 10-corollam fere aequantibus, filamentis liberis ligulatis 7-9 mm. longis ubique pallide puberulis, antheris circiter 40 mm. longis, thecis 9-11 mm. longis basi obtusis, tubulis quam thecis multo longioribus..gracillimis per poros obliquos circiter 0.5 mm. longos dehiscen- tibus; stylo filiformi leviter exserto, stigmate minuto. Cafiar: Near El Corazon, between S. Vicente and Rosario, 3,500 ft. elev., Sept. 6-10, 1944, F. Prieto CP-13 (TYPE US 1,988,905; dupl. NY) (high-growing epi- phyte; leaves dull on both surfaces and not markedly ‘‘veiny’’ when fresh; corolla pale rosy pink, the lobes bright green). Valley of rio de Cafiar at Abadel, below town of Galleturo, 6,000 ft. elev., Prieto CP-28 (epiphyte, the branches drooping, to 2 m. long; leaves dark green above, pale below, dull on both surfaces; corolla deep pink, the tips of lobes green); same locality, 4,400 ft. elev., Prieto CP-34 (epiphytic shrub; corolla crimson, the lobes green). This very distinct new species resembles G. charianthum in general aspect and like that species has a calyx-limb with minute teeth and flattened sinuses. It differs obviously in its smooth rather than winged ‘calyx-tube, its differently shaped and longer corolla with narrower lobes, its longer stamens, and its leaf- blades with the principal nerves oriented essentially from the base. Oreanthes glanduliferus A. C. Smith, sp. nov. Frutex epiphyticus, ramulis gracillimis obtuse angulatis subfuscis pallide puberulis et interdum parce glanduloso-hispidulis, mox glabrescentibus ciner- ascentibus; stipulis inconspicuis intrapetiolaribus oblongis obtusis circiter 1 mm. longis saepe ramulos adnatis; petiolis subteretibus rugulosis 1.5=3 mm. longis glanduloso-hispidulis (pilis 0.3-0.7 mm. longis) etiam puberulis mox glabratis; laminis in sicco coriaceis (in vivo subsucculentis et inflatis) elliptico-ovatis, 2.5-4.2 cm. longis, 1.3-2.3 cm. latis (apices ramulorum versus interdum 15 x 8 mm.), basi late obtusis vel subrotundatis, apice obtusis, margine siccitate sub- 64. MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol 8, No. 1 cartilagineis et paullo recurvatis, juventute utrinque ut petiolis parce glanduloso- hispidulis mox glabratis, e basi obscure 3- vel 5-nerviis, costa supra leviter im- pressa subtus elevata vel ut nervis utrinque immersa; inflorescentiis axillaribus 1-3-floris, rhachi subnulla, bracteis sub pedicellis deltoideis circiter 1 mm. longis; pedicellis teretibus gracillimis sub anthesi 5-20 mm. longis, pilis glanduloso- Capitatis 0.4-0.7 mm. longis copiose hispidulis ac etiam pilis eglandulosis circiter 0.1 mm. longis pallide hispidulis; calyce sub anthesi 7-10 mm. longo et circiter 3 mm. diametro cum pedicello continuo extus ut pedicello copiose glanduloso- hispidulo et puberulo, tubo obovoideo 3-6.5 mm. longo (paullo post anthesin), limbo erecto ad basim 5-lobato, lobis papyraceis lineari-subulatis 3.5-4,.5 mm. longis basi circiter 1 mm. latis intus glabris; disco carnoso annulri-pulvinato glabro; corolla in sicco submembranacea subcylindrica, 20-22 mm. longa, basim versus 3-5 mm. diametro, basi ipso et faucibus contracta, extus parce glanduloso- hispidula etiam pallide puberula, lobis sub anthesi patentibus oblongis 3.5-4 x 1-2 mm. apice obtusis; staminibus 5 corolla fere aequilongis, filamentis leviter cohaerentibus submembranaceis 2-2.5 mm. longis glabris, thecis levibus mem- branaceis circiter 3.5 mm. longis basi obtusis et leviter incurvatis, tubulis 14-15 mm. longis gracillimis per poros subterminales dehiscentibus; stylo filiformi leviter exserto, stigmate subtruncato, Cafiar: Valley of Rio de Cafiar at Abadel, below town of Galleturo, 6,000 ft. elev., Sept. 6-10, 1944, F. Prieto CP-29 (TYPE US 1,988,910; dupl. NY) (epiphyte; leaves pale green, only slightly shining, inflated and subsucculent; corolla crim- son); same locality, 4,400 ft. elev., .F. Prieto CP-37 (unicate, NY) (epiphyte; leaves subsucculent and inflated; corolla crimson). The second known species of Oreanthes, described above, differs from O. buxifolius Benth. in its conspicuous and slender pedicels, the copious hispid- glandular pubescence of its flowers (present also but less obvious on other parts of the plant), its comparatively short and essentially free filaments, and its differ- ently proportioned anthers, of which the thecae are much shorter. The genus Oreanthes remains an extremely rare entity in collections. It is apparently re- stricted to Ecuador, but since noting only the type collection of O. buxifolius in my 1932 treatment (Contr. U. S. Nat. Herb. 28: 359), I have seen two other collec- tions of that species, Penland 62 (NY) from Loja and 1180 (NY) from Tungurahua. Macleania macrantha Benth. Pl. Hartw. 223. 1846, Pichincha: Western slope of the cordillera, along the road from Quito to Sto. Domingo de los Colorados, about 6,000 ft. elev., Camp E-1737 (shrubs, arching to 2-3 m.; leaves very green above, very pale beneath; pedicels green; hypanthium winged, reddish or green; corolla deep crimson to coral-red, carnose; filament-tube pink, the connectives white, the anthers yellow). The type of this striking species also comes from the vicinity of Quito. It seems probable that Camp E-1718 (NY only) (shrub 1 m.; leaves deep green above, pale beneath; hypanthium grooved), from the same locality as no. 1737, represents a stage of M. macrantha with very young flower-buds. In this stage the corolla is subglobose and less than 5 mm. long, but the minute anthers are seen to be typically single-tubuled. A series of developing flowers would be very informative in eval- uating corolla-length as a specific criterion in this section of Macleania; it is likely that too much weight has been given to this character. Macleania floribunda Hook. Ic. Pl. 2: pl. 109. 1837. Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, Sierra de los Leones, near Bafios, 7,000 ft. elev., Camp E-1697 (NY only) (subepiphytic; flowers coral-red). 1952} PLANTS COLLECTED IN ECUADOR 65 The cited specimen agrees excellently in detail with the type and only previously known collection, Mathews 1442, from the Department of Amazonas, Peru. In no. 1697 the corolla has a fugacious white puberulence as well as the characteristic and more persistent brownish glandular hairs, this being the only difference noted between it and the type. Macleania recumbens A. C. Smith, sp. nov. Frutex subrecumbens, ramulis gracilibus subteretibus brunneis glabris mox de - corticantibus; petiolis rugulosis 3-5 mm. longis superne angulatis; laminis sub- coriaceis in sicco fusco-olivaceis ovatis, (3.5-)7-11 cm. longis, (2=)3-7 cm. latis, basi acutis et in petiolum decurrentibus, in acuminem ad 2 cm. longum gradatim angustatis raro tantum breviter acuminatis, margine integris leviter recurvatis, supra glabris, subtus dispersim et minute glanduloso-strigillosis, costa et nervis utrinsecus 2 vel 3 ad 2 cm. supra basim orientibus adscendentibus supra impressis subtus valde elevatis, venulis subimmersis; floribus axillaribus solitar- iis bracteis pluribus papyraceis deltoideis subacutis circiter 1 mm. longis subtentis, pedicellis gracilibus striatis (forsan 5-sulcatis) sub anthesi 8-9 mm. longis basim versus minute bibracteolatis; calyce turbinato sub anthesi circiter 9 mm. longo et apice 6 mm. diametro, tubo elongato circiter 7 mm. longo basim versus obscure pallido-glanduloso-strigilloso alis carnosis circiter 1.5 mm..latis manifeste S-alato, limbo erecto minutissime 5-denticulato, sinibus complanatis; corolla carnosa cylindrica circiter 25 mm. longa et basim versus 5 mm. diametro, faucibus contracta, intus apicem versus albido-pilosa alioqui glabra, lobis 5 oblongis sub- acutis circiter 3 mm. longis; staminibus 10 circiter 11 mm. longis, filamentis in tubum glabrum submembranaceum circiter 5 mm. longum connatis, antheris 7.5-8 mm. longis, thecis 4.5-5 mm. longis basi inflexis, tubulo unico 2.5-3 mm. longo conico, rima ovali subaequilonga; stylo filiformi corollam subaequante truncato. El Oro: In Moro-Moro region (about 21 miles west of Portovelo), 3,400-4,200 ft. elev., Oct. 7, 1944, Camp E-627 (TYPE US 1,988,986; dupl. NY) (plants subre- cumbent, with some branches to 1 m. long, in dense rain-forest; flowers solitary in axils of leaves; corolla deep coral-red). The new species is probably most closely related to M. floribunda Hook., dif- fering in its larger and longer-acuminate leaf-blades and its essentially glabrous (rather than distinctly castaneous-glandular-strigillose) flowers. The species in this section of Macleania (i. e. the species numbered 1 to 6 in my key in Contr. U. S. Nat. Herb. 28: 360. 1932) are fairly close and are not well represented in herbaria; an eventual reconsideration of specific lines will certainly be desira- ble. Within this alliance, the new species is close only to M. floribunda and M. angulata Hook., being distinguished from the latter by having its leaf-blades more definitely narrowed at both ends, its flowers solitary and with much shorter pedi- cels, and its corolla cylindric rather than angled. Macleania sleumeriana A. C. Smith, Contr. U. S. Nat. Herb. 29: 367. 1950. Pichincha: Western slope of the cordillera, along the road from Quito to Sto. Domingode los Colorados, about 6,000 ft. elev., Camp E-1735 (NY only) (terrestrial shrubs 0.4-1 m. high; leaves deep green above, pale beneath; upper part of pedi- cels and hypanthium bright red-orange at anthesis, later fading to cream-yellow, with only the wings red-tinged; base of corolla reddish coral, the upper half deep green). The species (Anthopterus ericae Sleumer, non Macleania ericae Sleumer) appears to be rare; this is only the third collection known to me, all being from northern Ecuador. | , . 66 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 Macleania salapa (Benth.) Benth. & Hook. Gen. Pl. 2: 566; 1876. Loja: Nudo de Cajanuma (south of Loja), 7,600 ft. elev., Camp E-569 (shrub 2 m.; hypanthium ribbed; corolla pale pink). Mataperro region, a pass between the Cordillera de Cisne and the Cordillera Chicuanco, about halfway between Zaruma and Loja, 7,100 ft. elev., Camp E-646 (shrub to 3 m., common on dry soil in the pass; flowers pale red). Apparently limited to southern Ecuador; I have seen specimens from only Loja and El Oro, The combination was not properly made in Genera Plantarum, and per- haps the correct citation of the authority should be (Benth.) Hook. f. ex Hoer. Bot. Jahrb. 42: 269. 1909. Macleania rupestris (H. B. K.) A. C. Smith, Phytologia 1: 131. 1935. Carchi: Slopes of Volcan Chiles, Camp E-319, E-320, E-327, E-334. East of Tulcan, Camp E-363. Pichincha: West of Quito, on Sto. Domingo Road, Camp E-1706, E-1707, E-1719. Cafiar: Between Tambo and Suscal, north rim of the valley of the Rio de Cafiar, Camp (coll. M. Giler) E-2766, E-2770A, E-2770B. Azuay: Along the Rio Matadero, west of Cuenca, Camp E-1942, E-1983, E-1985. Valley of the Rio Surucuchu, west of-Cuenca, Camp E-4240. Along the Rio Cumbe, south of Cuenca, Camp E-2077, E-2080. Paramo de Tinajillas, south of Cuenca, Camp E-481. Loja: Cerro Villanaco, west of Loja, Camp E-233., Dr. Camp’s notes pertaining to the extensive suite of specimens cited above are very detailed. Briefly summarized, they indicate that the species was observed in Ecuador at elevations of 8,500-11,400 ft., occurring on paramo, in the paramo- sotobosque zone, or in subparamo chaparral; specimens were noted as small trees (rarely) or more often as spreading or sprawling shrubs 1.5-6 m. high, sometimes as much-branched vines climbing through low trees; soft-tissued basal burls were often observed; the corolla is pale crimson to pink at base and paler or white distally; the mature fruit is as much as 1,5 cm. in diameter, shining black, and insipid. In 1932 (Contr. U. S. Nat. Herb. 28: 360-384) I recognized ten Ecuadorian species in the group of Macleania with 2-tubuled anthers, although some of these were admittedly segregated on rather insignificant characters. In addition, four species based on Ecuadorian types were reduced to synonymy; one other species from Ecuador, M. mollis, has been described more recently. The accumulation of herbarium material since 1932 and a study of the present material incline me to believe that specific lines in the 2-tubuled Macleaniae cannot be satisfactorily ~ established by observational methods. Particularly in Ecuador, which seems to be a center of development of the group, the usual specific criteria are combined in — such diverse ways that one must assume free inter-breeding among the ‘‘species’’ to be a continuing phenomenon. In view of this, to apply specific names to parts — of the population is perhaps undesirable; but nevertheless I have identified the Camp collections according to current concepts, with the reservation that these concepts may be far from natural. Macleania rupestris (based on the oldest available specific epithet for this group, Thibaudia rupestris H. B. K. 1818), in the strict sense, is characterized by being essentially glabrous throughout, with flowers of moderate size (corolla usually 15-20 mm. long) and stamens with tubules subequaling the thecae in length. Its leaves are variable, but in general they are rounded to acute at base and pinnatinerved. The Central American M. glabra (K1.) Hoer. is scarcely t to be distinguished from the common South American species. Macleania pilgeriana Hoer. Bot. Jahrb, 42: 301. 1909. Pichincha: Paramo west of Quito on Sto. Domingo Road, 11,300 ft. oh Camp E-1705, E-1710 (shrubs 1=2 m.; corolla deep red). 1952] PLANTS COLLECTED IN ECUADOR 67 The cited specimens agree well with type material (also from Pichincha) in foliage and in the elongate anther-tubules; I doubt whether this entity should be kept apart from M. rupestris even ona subspecific level. Macleania benthamiana Walp. Repert. Bot. 6: 415. 1847. Chimborazo-Cafiar border: Western escarpment, between Sta. Rosa and Joyagshi, 8,300-9,000 ft. elev., Camp E-4043 (shrub to 2 m., from relatively small burl). Azuay: Paramo de Tinajillas and surrounding chaparral and forests, 30-50 km. south of Cuenca, 9,200 ft. elev., Camp E-453 (spreading shrub to 3 m.), Same locality, 11,000-11,500 ft. elev., Camp E-2285 (shrub 2 m., from burl about 0.3 m. in diameter). The cited specimens have leaves which are deep green above and paler be- neath; the hypanthium is greenish to deep red, the corolla crimson to coral or pale pink toward base, paler distally, the filaments white, the anthers orange. The numerous oblong-lanceolate bracts subtending the inflorescence distinguish this entity from M. rupestris. It is essentially glabrous throughout and typically has rather large and coriaceous leaves, with the principal nerves strongly raised beneath; the cited specimens, however, have leaves smaller than typical. Macleania ecuadorensis Hoer. Bot. Jahrb. 42: 300. 1909. Cafiar: Uplands called ‘‘Huairacaja,’’ 10-20 km. northeast of Azogues, 11,000 ft. elev., Camp E-1758 (shrub 4 m.; leaves dull, deep green above, pale beneath; corolla deep pink toward base, apically white, becoming crimson with age). Azuay: Paramo del Castillo and surrounding forested areas (crest of the eastern cordillera on the trail between Sevilla de Oro and Mendez), 9,000-11,000 ft. elev., Camp E-725A (NY only) (erect shrub 2 m.; flowers deep red). Paramo and sub- paramo area north and northwest of the Paramo del Castillo, 10,000-11,200 ft. elev., Camp E-5157 (NY only) (shrub 2 m.; leaves deep green above, pale beneath; corolla basally crimson, the apex and lobes pale pink). Macleania ecuadorensis has the leaves characteristically white-pilose beneath, with prominent secondary nerves, and regularly oval in shape. However, no. 1758 has narrower than typical leaves, while nos. 725A and 5157 have the calyx and corolla faintly pilose. The entity seems hardly more than an expression of character- combinations in the general complex of M. rupestris and M. hirtiflora. Macleania loeseneriana Hoer. Bot. Jahrb. 42: 302. 1909. Pichincha: Western slope of the cordillera, Cerro Corazon, 11,000 ft. elev., Camp E-1647 (NY only) (spreading shrub to 3 m.; leaves dull on both surfaces, deep green above, paler beneath; hypanthium red; corolla basally pale red, apically white; immature fruit dull, non-glaucous). The robust habit and inflorescence, pilose flowers, and calyx with large sub- spreading limb make this one of the more easily identified entities among the 2-tubuled Macleaniae. Macleania hirtiflora (Benth.) A. C. Smith, Contr. U. S. Nat. Herb. 28: 382. 1932. Cafiar: Uplands called ‘*‘Huairacaja,’’ 10-20 km. northeast of Azogues, Camp E-1756. Azuay: Paramo de Tinajillas and surrounding chaparra! and forests, 30-50 km. south of Cuenca, Camp E-386. Cordillera de Alpachaca, Camp E-285, E-532, E-536, s. n. (May 22, 1944), Paramo de Carboncilla, about 15 km. south of Ofia, Camp E-554A-E-544F incl. (NY only). Paramo del Castillo and surrounding forested areas (crest of the eastern cor@illera on the trail between Sevilla de Oro and Mendez), Camp E-4844. Eastern Cordillera, between Ofia and the Rio Yacuambi, F, Prieto P-23] (NY only). | The cited specimens were collected at elevations of 8,000 to 11,200 ft., on paramo or subparamo; they are noted as shrubs up to 4 m. high, often spreading meee 2 ef na ee oe = — =< . ah 7 68 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 from large burls (these sometimes more than 1 m. in diameter); leaves deep green and subnitid above, paler beneath; hypanthium often red or crimson; corolla deep crimson to pink at base, paler or yellowish or white distally; fruit elongate-spherical, when ripe purple-black and up to 1.5 cm. in diameter, sweetish or insipid. These specimens are variable in details, as indeed are those I referred to this species in 1932 and the numerous ones so identified since that time. It seems unlikely that this concept represents a natural genetic unit; actually the specimens might be construed as representing extreme forms of various strains of M. rupestris, characterized by the pilose flowers, a tendency toward shorter corollas and stamens, and frequently pilose leaves. The indument seems persistent on the pedicels and calyx, but it is sometimes fugacious on the corolla. Among the specimens from Azuay, nos. 285, 386, 532, and 536 have an unusually persistent corolla-indument, as well as a tendency toward very small leaves. Macleania cf. hirtiflora (Benth.) A. C. Smith. Cafiar: Paramo between Biblian and Cafiar, Camp E-447, Parroquia Bayas, valley of Rio Tabacal, about 15 km. northeast of Azogues, F. Prieto P-118, Region of San Marcos, about 10 km. northeast of Azogues, F, Prieto P-80, Uplands called “‘Huairacaja,’’ 10-20 km. northeast of Azogues, Camp E-1754, E-1811, Azuay: Cruz Pamba region above Bafios, about 15 km. southwest of Cuenca, Camp E-3939 (coll. M. Giler & F. Prieto). The eastern Cordillera, 4-6 km. north of the village of Sevilla de Oro, Camp E-4702. Vicinity of El Pan, Camp E-500, Paramo del Castillo and surrounding forested areas (crest of the eastern cordillera on the trail be- tween Sevilla de Oro and Mendez), Camp E-716, E-721 (NY only), E-724 (NY only). The cited specimens were obtained at elevations between 8,500 and 11,000 ft.; they came from erect, spreading, or scrambling shrubs 2=5 m. high, sometimes with large soft burls; leaves deep green and subnitid above, pale and dull be- neath; hypanthium crimson; corolla crimson to pink, tipped with yellow, white, or pale pink; fruit ellipsoid, when mature about 2 x 1.5 cm.,°pink- or red-flushed, sweetish but flat in taste; local name guayapa, used for plants of this general affinity. These collections agree with M. hirtiflora in the indument of their flowers and sometimes of their foliage, but their inflorescences are subtended by elongate bracts similar to (or approaching in size) those of M. benthamiana. It must be assumed, I think, that these two species are interfertile where their ranges coin- cide, if indeed either species is more than a series of variations from M. rupestris. The cited specimens also suggest other ‘‘species’’ of this complex. In some, e. g. no. 4702, the leaves may be coriaceous‘and prominently nerved as in typical M. benthamiana, but sometimes strictly glabrous and sometimes pilose in precisely the manner typical for M. ecuadorensis. Number 721 is accompanied by extraordi- narily large leaves (blades up to 16x 10 cm.), although its inflorescences are associated with leaves of normal size. Macleania mollis A. C. Smith, Phytologia 1: 132. 1935. Cafiar: Valley of Rio de Cafiar at “‘Selem,’’ between Galleturo and Cafiar, 7,000 ft. elev., F. Prieto CP-38, CP-39 (terrestrial shrubs 2 m.; leaves deep mag dull or somewhat shining; corolla coral-red). | These are the only collections of the species known to me except for the type, from Chimborazo. While this entity is not too distinct from certain forms of M. hirtiflora, it is distinguishable by its subcordate leaf-blades with more basally oriented and ascending secondary nerves, its few-flowered fasciculate inflores- cences, and its corollas averaging longer. On the whole, M. mollis seems a stronger ‘‘species’’ than most of this relationship. 1952] PLANTS COLLECTED IN ECUADOR 69 Macleania coccoloboides A. C. Smith, sp. nov. Frutex ad 3 m. altus vel epiphyticus et subscandens ubique praeter ramulos et petiolos glaber, ramulis robustis subteretibus nigrescentibus primo et petiolis minute pallido-puberulis mox glabratis; petiolis crassis (2-5 mm. diametro) sub- teretibus rugulosis 3-7 mm. longis; laminis in sicco coriaceis fusco-viridibus suborbiculari-ovatis, (4=)6=15 cm. longis, (3-)4-9.5 cm. latis, basi rotundatis vel leviter cordatis, apice rotundatis (juventute forsan late obtusis), margine incrassa- tis et leviter recurvatis, ubigque dispersim punctato-glandulosis, pinnatinerviis, costa supra leviter insculpta subtus prominente, nervis secundariis utrinsecus 3 vel 4 supra subplanis vel prominulis subtus valde elevatis, inferioribus e costa basim versus orientibus curvato-adscendentibus, superioribus debilioribus, rete venularum immerso vel utrinque haud prominulo; inflorescentia axillari congesta breviter racemosa ut videtur 7=10-flora bracteis numerosis subcoriaceis oblongo- deltoideis circiter 2 mm. longis basi circumdata, bracteis floriferis reniformibus 1-1.5 mm. longis, rhachi crassa sub fructu 5—7 mm. longa; pedicellis crassis teretibus rugulosis sub anthesi ad 9 mm. sub fructu ad 20 mm. longis, apice con- Spicue incrassatis et margine apicali obscure glandulosis, cum calyce manifeste articulatis, basim versus bibracteolatis, bracteolis late deltoideis 1-2 mm. longis circiter 2.5 mm. latis obtusis; calyce sub anthesi 7-8 mm. longo et apice diametro ubique dispersim nigro-punctato-glanduloso, tubo cupuliformi circiter 5 mm. longo, limbo suberecto carnoso 2-3 mm. longo minute 5-denticulato, sinibus complanatis; corolla carnosa urceolato-cylindrica sub anthesi 13-14 mm. longa et basim versus circiter 7 mm. diametro, superne paullo angustata, lobis 5 deltoideis subacutis circiter 1.5 mm. longis; staminibus 10 quam corolla multo brevioribus, filamentis ligulatis 1.5-2 mm. longis, connectivis latis, antheris circiter 7 mm. longis, thecis crassis quadrangularibus circiter 5 mm. longis basi obtusis et incurvis in tubulos 2 graciles acutos circiter 2 mm. longos saepe ad basim liberos rimis elongatis ovalibus dehiscentes abrupte angustatis; stylo filiformi corollam subaequante, stigmate minuto; fructibus elongato-subglobosis ad 1 cm. diametro calycis limbo coronatis. Pichincha: Westem slope of the cordillera, along the road from Quito to Sto. Domingo de los Colorados, 7,000-8,500 ft. elev., Jan. 15, 1945, Camp E-1726A (TYPE US 1,989,052; dupl. NY) (shrub 1-3 m., arching from banks, arising from a burl up to 0.3 m. in diameter, also seen as a high-growing epiphyte and then also with burl; branches sooty-black; leaves very deep green above, pale to subglaucous beneath, with glands at first black and later conspicuous by the presence of a white fungus; hypanthium pale yellow to red, very shallowly grooved, the calyx margin and sometimes the base with notable pits; corolla deep crimson; fruit rip- ening deep purple-black, shining), E-1726B (NY only) (young plant of no. 1726A, showing the burl, this soft-parenchymatous, not woody). Same general locality, Cerro Corazén, 8,000-9,300 ft. elev., Camp E-1679 (NY only) (epiphytic and vine- like; twigs dark, blackish; leaves exceptionally dark green above, pale and glaucous beneath; immature fruits pale, subtranslucent). This new species is readily distinguished by its comparatively short corollas | and anthers abruptly terminating in very short, slender tubules. It does not seem to have very close allies, but in some ways it suggests M. costeroides Sleumer (Bot. Jahrb. 71: 401. 1941), also of Ecuador. From this, M. coccoloboides differs in its larger leaf-blades with more highly connate nerves, its very short inflores- cences, minutely denticulate calyx-limb, longer corolla, and stamens with free filaments and separate tubules. Psammisia corallina A. C. Smith, sp. nov. Frutex epiphyticus ubique glaber; ramulis robustis obtuse angulatis in sicco Striatis fuscis; petiolis valde rugulosis angulatis incrassatis' 10-12. mm. longis; \. 70 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 laminis coriaceis siccitate fuscis ellipticis, 10-15 cm. longis, 6-7.5 cm. latis, basi acutis et in petiolum conspicue decurrentibus, apice breviter et obtuse cuspi- datis, margine valde recurvatis, 5-nerviis, nervis summis cum costa 1-3 cm. concurrentibus et cum costa supra leviter impressis (basim versus elevatis) subtus prominentibus, nervis infimis basalibus inconspicuis, rete venularum utrinque prominulo vel immerso; inflorescentiis 1-3 in axillis foliorum breviter racemosis 6-15-floris, floribus caducis, rhachi robusta 2-4 cm. longa articulationibus valde incrassata, bracteis sub floribus papyraceis deltoideis 1=1.5 mm. longis subacutis; pedicellis robustis sub anthesi 15-25 mm. sub fructu juvenili ad 35 mm. longis basim versus bibracteolatis, bracteolis minutis; calyce carnoso sub anthesi 6-7 mm. longo et 8-9 mm. diametro, tubo obtuse angulato 3~4 mm. longo, limbo sub- patente 5-denticulato margine leviter incrassato et inflexo, lobis haud 1.5 mm. longis apiculatis, sinibus complanatis vel rotundatis; disco conspicue annulari- pulvinato; corolla carnosa subgloboso-conica sub anthesi circiter 9 mm. longa et 8 mm. diametro, lobis deltoideis subacutis circiter 1.5 mm. longis; staminibus 10, filamentis in tubum 1.5-2 mm. longum connatis, Connectivis ecalcaratis, antheris maturis circiter 7.5 mm. longis, thecis crassis circiter 6 mm. longis basi obtusis supeme in tubulos breves acutos cum rimis angustis angustatis; stylo tereti corollam subaequante truncato. Santiago-Zamora: Cordillera. Cutuci, ridge between the Rios Itzintza and Chupiasa, 4,000-4,500 ft. eley., Nov. 17-Dec. 5, 1944, Camp E-1280 (TYPE US 1,989,014; dupl. NY) (epiphyte, with branches about 1 m.; leaves dark green above, pale beneath; pedicels and hypanthium pale pink, the corolla coral-red). From its closest relative, the Colombian P. occidentalis A. C. Smith, the new species is readily distinguished by its much more robust inflorescence, with stouter and longer rachis and pedicels and larger flowers; the largest corollas seen in P. occidentalis hardly exceed 5 mm. nor the longest anthers 3 mm. in length. The Ecuadorian P. flaviflora A. C. Smith (Jour. Arnold Arb. 24: 463, 1943) is less closely related to the new species, differing in its caudate-acuminate leaf- blades, short inflorescence, calyx-limb with larger lobes and acute sinusés, free filaments, and shorter, spurred anthers. Psammisia ecuadorensis Hoer. Bot. Jahrb. 42: 308. 1909. Cafiar: Near El Corazon, between S. Vicente and Rosario, 3,500 ft. elev., F. Prieto CP-12 (low-growing epiphyte with pendant branches 1-3 m. long; leaves dull above, shining beneath; pedicels coral-red; corolla deep crimson). The species is known from several Ecuadorian collections and possibly ex- tends gorthward into Colombia. Psammisia_ ferruginea A. C. Smith, Contr. U. S. Nat. Herb. 28: 391. pl. 10. 1932. Pichincha: Western slope of the cordillera, along the road from Quito to Sto. Domingo de los Colorados, about 6,000 ft. elev., Camp E-1736 (shrub, with branches 3m. long). Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, Shell Mera (east of Mera), about 3,500 fr. elev., Camp E-1701 (NY only) (large spread- ing plant, in part epiphytic; flowers pink). Same locality, uplands near El Topo, along trail to La Gloria, 4,000-5,000 ft. elev., Camp E-2399 (NY only) (arching epiphyte, common at lower elevations). Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), between Tres Ranchos and Chontal, 2,700-5,700 ft. elev., Camp. E-1566 (epiphytic vine; hypanthium pale red; corolla pale yellow at anthesis, later pale red). Cordillera Cutuci, ridge ascending into central Cutuct, 4,400-4,700 fr. elev., Camp E-1159 (immense climbing and epiphytic plants, some 5 m. across and hanging over 10 m., common along streams on west slope of the Cutuci; pedicels 1952 | PLANTS COLLECTED IN ECUADOR aL and calyx pink-tinged; corolla yellow in bud, at anthesis tinged with pink under the crimson hairs; body of corolla red at about time it is ready to fall, and crimson as it lies on the ground; some plants with flowers redder than others at anthesis). Same locality, on banks of Rio Itzintza, 3,500 ft. elev., Camp E-1199 (high-climbing epiphyte, common in this region but seen only along streams; leaves green above, paler beneath; pedicels and hypanthium crimson; corolla yellow in bud, with crimson hairs, becoming red or even crimson in age, apically constricted; filaments and connectives white, the anthers yellow). The excellent series of specimens cited forms a welcome addition to the herbarium material of this species, which otherwise I have known only from southern Colombia (Cauca, El Valle, Narifio, and Putumayo). Dr. Camp’s material, of course, demon- strates a few minor variations from the original description, but the fundamental characters of the species are unmistakable. Psammisia sodiroi Hoer. Bot. Jahrb. 42: 306. 1909. Pichincha: Western slope of the cordillera, Cetro Corazén, 8,000-9,300 ft. elev., Camp E-1680 (vine-like epiphyte; leaves pale green and dull above, paler but subnitid beneath; pedicel basally greenish, apically bright coral-red; hypan- thium bright coral-red at anthesis, the color fading to dark green as the fruit enlarges; corolla deep red toward base, pale green in upper half; flowering irreg- ular, sometimes at apex of stem, or later on nearly bare wood). Along the road from Quito to Sto. Domingo de los Colorados, 7,000-8,500 ft. elev., Camp E-1728 (NY only) (shrub 3 m.; leaves pale green and dull above, subnitid beneath; hypanthium pale coral-red; base of corolla deep crimson, the apex green). The cited specimens are very typical of the species, which is known from several collections from Pichincha and extends northward into Narifio. Sleumer inadvertently omitted this species from his review of the Psammisiae with pinnati- nerved leaves (Bot. Jahrb. 71: 403-404. 1941), Psammisia oreogenes Sleumer, Bot. Jahrb. 71: 403. 1941. Pichincha: Western slope of the cordillera, Cerro Corazon, 8,000-9,300 fe. elev., Camp E-1677 (vine-like epiphyte; leaves deep green above, pale beneath; at anthesis peduncles deep coral-red and hypanthium pale coral, deeply grooved; base of corolla red, white toward the end, the apex green; bracts and pedicels green; flowers in axils of leaves), Camp E-1678 (NY only) (epiphytic vine, with flowers on old wood; pedicels deep coral-red, the bracts green; hypanthium pale yellowish coral; lower part of corolla red, the apex green, without white zone noted in no. 1677 but in same stage, the corolla much broader). In foliage these two specimens appear conspecific, but unfortunately the flowers described for no. 1678 have been lost. Differences can be observed between these plants and the description of P. oreogenes, typified by Heilborn 488, also from Pichincha. Dr. Camp’s specimens have the petioles slightly longer, the pedi- cels shorter, more slender, and glandular-pilose distally rather than glabrous, the calyx glandular-strigillose rather than glabrous, the stamens somewhat longer (about 9 mm. long), and the anther-tubules about 5 mm. rather than 2.5=3 mm. long, - Our specimens differ from the allied P. sodiroi in the merely apiculate calyx- limb, the longer corolla, and in details of venation, in which characters they seem to agree with Sleumer’s species. Psammisia idalima A. C. Smith, sp. nov. Frutex subscandens ubique praeter filamenta glaber, ramulis gracilibus sub- teretibus fuscis, intemodiis bracteas papyraceas lanceolato-oblongas 5-10 mm. longas obtusas interdum gerentibus; petiolis crassis subteretibus rugulosis ni- ‘ Te MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 grescentibus 5-10 mm. longis; laminis in sicco subcoriaceis metallico-olivaceis oblongo-ellipticis, (12-)15=-25 cm. longis, (3.5=)5-10 cm. latis, basi late obtusis et in petiolum subito decurrentibus, apice breviter acuminatis, margine leviter recurvatis, pinnatinerviis, costa supra elevata vel superne impressa subtus promi- nente, nervis lateralibus utrinsecus 5=7 arcuato-adscendentibus anastomosantibus supra subplanis subtus elevatis, rete venularum subimmerso vel utrinque paullo prominulo; inflorescentia axillari vel infra folia enata breviter racemosa 6-9-flora; rhachi angulata 5-10 mm. longa, bracteis sub floribus papyraceis oblongis 2=3 mm. longis obtusis; pedicellis gracilibus sub anthesi 12-18 mm. sub fructu ad 20 mm. longis basim versus bibracteolatis, bracteolis subpapyraceis ovato-deltoideis 2=2.5 mm. longis circiter 1.5 mm. latis subacutis pauciglanduloso-marginatis; calyce sub anthesi circiter 6 mm. longo et apice 7~8 mm. diametro, tubo cupuliformi circiter 3 mm. longo, limbo papyraceo erecto-patente profunde 5-lobato, lobis oblongo- ovatis 2-3 mm. longis circiter 3 mm. latis praeter apicem apiculatum crasso- marginatis, sinibus acutis; corolla camosa urceolata sub anthesi 7.5-8 mm. longa et basim versus circiter 5 mm. diametro faucibus angustata, lobis 5 oblongis subacutis circiter 1 mm. longis; staminibus 10, filamentis submembranaceis circi- ter 1mm. longis interdum superme intus minutissime pilosis in connectivos graciles ecalcaratos angustatis, antheris 3.5=4 mm. longis, thecis circiter 2.5 mm. longis basi incurvatis, tubulis 1-1.5 mm. longis per rimas elongatas dehiscentibus; stylo tereti corollam subaequante, stigmate minuto; fructibus subglobosis in sicco coriaceis rugulosis ad 1 cm. diametro calycis limbo persistente et disco cori- aceo pulvinato ad medium depresso coronatis. Azuay: The eastem Cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000-9,000 ft. elev., July 27=Aug. 12, 1945, Camp E-4597 (TYPE US 1,989,098; dupl. NY) (plant vine-like; leaves deep green above, very pale beneath; hypanthium crimson; base of corolla crimson, the apex white; fruit non-glaucous), Camp E-4379 (vine; leaves deep green and dull above, pale green beneath; pedicels and hypanthium crimson; corolla basally deep pink, apically pale pink to white; imma- ture fruit nitid). From P. sodiroi Hoer., apparently its closest ally, the new species differs; in having its leaf-blades slightly thicker in texture and with less prominent venation, its pedicels and calyx more slender, its corolla shorter, and its anthers much shorter and essentially ecalcarate. The small flowers and other obvious combina- tions of characters readily separate P. idalima from other species of this immediate relationship, P. graebneriana Hoer., P. debilis Sleumer, and P. oreogenes Sleumer. Psammisia sclerantha A. C. Smith, sp. nov. Frutex parvus interdum epiphyticus ubique praeter filamenta glaber, ramulis teretibus gracillimis fusco-stramineis; petiolis inconspicue angulatis 7-15 mm. longis; laminis subcoriaceis in sicco metallico-olivaceis elliptico-lanceolatis, 11-21 cm. longis, 4-6.5 cm. latis, basim versus gradatim angustatis et in petio- lum decurrentibus, apice longe et acute acuminatis, margine leviter recurvatis, pinnatinerviis, costa supra paullo subtus valde elevata, nervis lateralibus utrin- secus 4 vel 5 erecto-patentibus anastomosantibus supra subplanis subtus paullo elevatis, venulis utrinque haud prominulis; inflorescentia axillari breviter race- ‘mosa 2=7-flora; rhachi gracili obtuse angulata ad 12 mm. longa, bracteis sub floribus papyraceis deltoideo-oblongis obtusis 1-1.5 mm. longis latisque; pedi- cellis sub anthesi 13-20 mm. longis superne incrassatis et sub calyce conspicue articulatis basim versus bibracteolatis, bracteolis bracteis similibus minoribus; calyce cupuliformi sub anthesi circiter 6 mm. longo et 8 mm. diametro, tubo brevi et lato, limbo crasso-carnoso erecto-patente 3-4 mm. longo lobis haud 1 mm. longis inconspicue 5-dentato praeter apices loborum obscure crasso-marginato, i 1952] PLANTS COLLECTED IN ECUADOR 73 sinibus complanatis vel late obtusis; corolla cylindrica sub anthesi 8-9 mm. longa 4—6 mm. diametro superne conspicue crasso-carnosa, lobis 5 deltoideis circiter 2 x 2 mm. acutis sub anthesi inflexis; staminibus 10, filamentis submem- branaceis pallidis ligulatis 2~3 mm. longis superne intus obscure puberulis in connectivos latos omnibus obtuse sed manifeste calcaratos transeuntibus, antheris 4,5—6 mm. longis, thecis crassis subquadratis 3.5-4.5 mm. longis basi incurvatis et obtusis in tubulos graciles 1-1.5 mm. longos cum rimis elongatis ovalibus abrupte angustatis; stylo tereti corollam fere aequante stigmate obscure lobato; fructibus juvenilibus late subglobosis ad 1 cm. latis calycis limbo persistente inflexo et disco coriaceo pulvinato coronatis. Santiago-Zamora: Cordillera Cutucu, ridge between Rio Ontza and Rio Chupiasa, 4,300~4,700 ft. elev., Nov. 17=Dec. 5, 1944, Camp E-1195 (TYPE US 1,989,009; dupl. NY) (leaves deep green above, pale green beneath, dull on both surfaces; pedicels and hypanthium coral-red; corolla green in bud, red as it begins to open, and purple with age). Same locality and altitude, ridge ascending into central Cutuct, Camp E-1156 (small shrubs 1 m., epiphytic or terrestrial; pedicels and calyx deep coral-red; corolla green, carnose, very hard when open; plants later seen in shade, with nearly white flowers). Psammisia sclerantha is another of the species with pinnatinerved leaves, characterized by having the calyx-limb and distal part of its corolla extraordinar- ily thick in texture and its anthers with broad, spurred connectives and very short, slender tubules. From P. sodiroi Hoer. and the other species of this alliance except P. oreogenes Sleumer, the new species differs in its merely denticulate (rather than conspicuously lobed) calyx-limb. Psammisia oreogenes, however, has comparatively short-petioled leaves and a differently proportioned calyx (the limb being shorter and thinner) and anthers. Psammisia columbiensis Hoer. Bot. Jahrb. 42: 303. 1909, Azuay: The eastern Cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000~9,000 ft. elev., Camp E-4470 (large vine, scrambling to 7 m.; leaves deep green and dull above, pale and subnitid beneath; peduncles bright green; pedicels red at base, becoming crimson above; hypanthium dull crimson, the calyx-lobes tipped with yellow; corolla doubly constricted, crimson to second constriction, the apex and lobes white; filaments united at base; immature fruit dull green). I am unable to distinguish the cited specimen from P. columbiensis, typified by a specimen from Cauca and also now known from Antioquia and Putumayo in Colombia. The species is characterized by its narrow, few-nerved leaves, its elongate inflorescence, and its flowers of medium size for the genus, with large calyx-lobes and connate filaments. The Camp collection has the filaments only loosely united and the anthers very inconspicuously spurred, these points of dif- ference from typical material being the only ones observed. Another specimen that should be considered here is: Camp E-1144 (Santiago- Zamora: Cordillera Cutuci, ridge ascending into central Cutuct, 4,400-4,700 ft. elev., a high-climbing often epiphytic vine; pedicels, calyx, and base of corolla deep coral-red, the apex of corolla white). This specimen differs from no. 4470 and Colombian material of the species in having its leaf-blades thinner in texture and with more obvious venation, and in having its anthers only about 7 mm. (rather than 10-11 mm.) long; the filaments are clearly connate in some flowers and essentially free in others, while the inflorescence is characteristically elongate. Until a more comprehensive suite of specimens of this immediate alliance is available, I hesitate to suggest that more than one species is included, although this may prove to be the case. ‘ + 74 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 Psammisia guianensis Kl]. Linnaea 24: 43, 1851. Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, Shell Mera (east of Mera), about 3,500 ft. elev., Camp E-1703 (in swampy areas, often epi- phytic, with arching branches to 3 m. long; inflorescences coral-red, the young corolla white). Santiago-Zamora: Valley of the Rio Zamora, east of Loja, ridge across river from the village of Zamora, about 6,500 ft. elev., Camp E-31 (NY only) (shrub sprawling over rocks). Uplands along Rio Upano just north of junction with Rio Chupianza, near Mendez, 1,750-2,500 ft. elev., Camp E-1004 (NY only) (shrubs on ground or on rotting logs, arching to 3 m.). Eastern slope of the cor- dillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), El Partidero, between the Rios Paute and Negro, 2,100-3,100 ft. elev., Camp E-1522 (large epiphytic vine, the branches to 6 m.; leaves deep green above, pale beneath; pedicel, hypanthium, and lower part of corolla deep coral-red, the tip of corolla above constricted part white; filaments and connectives white; anthers brown; immature fruit non-glaucous),. The cited specimens agree excellently with other material representing this widespread species from the eastern slopes of the Andes. Psammisia ulbrichiana Hoer. Bot. Jahrb. 42: 306. 1909. Azuay: The eastern cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000-9,000 ft. elev., Camp E-43067 (spreading shrub, the branches ultimately vine-like, to 5 m. long; leaves dark green and subnitid above, pale beneath; im- mature fruit pale salmon), Camp E-4402 (vine; leaves deep green and nitid above, pale beneath; immature fruit pale salmon). Santiago-Zamora: Cordillera Cutuci, ridge ascending into central Cutuct, 3,500 ft. elev., Camp E-1102 (NY only) (high- growing epiphyte, the branches arched to 2 m.; leaves deep green above, pale be- neath). Same locality, ridge between Rios Itzintza and Chupiasa, 4,000-4,500 ft. elev., Camp E-1271 (coarse climbing epiphyte, with branches to 5 m. long). The cited specimens are all in fruit, but the comparatively large leaves, short inflorescences with congested floral scars, and calycine characters point to their position in P. ulbrichiana, typified by a specimen from the Province of Pichincha. Psammisia sp. Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, Rio Tigre, near junction with the Pastaza (below Topo), 5,400 ft. elev., Camp E-1693 (epi- phyte with branches 5 m. long). In the shape and texture of its leaves, the cited specimen suggests P. pauciflora Griseb. ex A. C. Smith, but its persistent calyx-lobes in fruit are rather large for that sbecies. Flowers are needed satisfactorily to place the specimen. Calopteryx sessiliflora A. C. Smith, sp. nov. Frutex, ramulis fuscis obtuse angulatis pallide puberulis mox glabratis; stipu- lis intrapetiolaribus lanceolato-subulatis circiter 7 mm. longis caducis; laminis e tamulis brevibus lateralibus interdum orientibus, ramulis bracteis papyraceis lineari-lanceolatis ad 15 x .2 mm. circumdatis; petiolis incrassatis subteretibus subglabris 5-12 mm. longis; laminis papyraceis in sicco fusco-olivaceis anguste lanceolatis, 13-28 cm. longis, 3.5-7.5 cm. latis, basi late obtusis, in apicem 1-2 cm. longum gradatim angustatis, subtus pilis glandulosis fusco-castaneis 0.2-0.3 mm. longis copiose strigillosis, supra mox glabratis, 5-7-nerviis, nervis — secundariis adscendentibus cum costa 2=7 cm. concurrentibus ut costa supra im- pressis (vel basim versus leviter elevatis) subtus prominentibus, rete venularum supra subimmerso subtus prominulo; inflorescentia ramulis defoliatis enata sub- fasciculata ut videtur l- vel 2-flora bracteis numerosis circumdata, bracteis lineari- — 1952] PLANTS COLLECTED IN ECUADOR a lanceolatis obscure glanduloso-marginatis, majoribus circiter 15 x 2 mm., intimis calycem involventibus circiter 10 x 3 mm.; pedicellis subnullis, bracteolis 1 vel 2 lineari-subulatis circiter 7 x 0.5 mm.; calyce 6.5=7 mm. longo et circiter 4 mm. apice diametro pilis minutis glandulosis strigilloso, tubo obtuse 5-angulato circiter 4.5 mm. longo, limbo quam tubo breviore intus glabro fere ad basim 5-lobato, lobis elongato-deltoideis circiter 3 x 2 mm.; disco annulari-pulvinato glabro; corolla tenuiter carnosa parce glanduloso-strigillosa circiter 13 mm. longa et 4 mm. diametro anguste 5-alata, alis medio circiter 0.7 mm. latis superne angustatis, lobis deltoideis subacutis circiter 1 mm. longis; staminibus 10 corollam sub- aequantibus, filamentis membranaceis circiter 2.5 mm. longis ut videtur connatis superne obscure puberulo-marginatis, antheris circiter 10.5 mm. longis, thecis 4~4.5 mm. longis basi mucronatis, tubulis quam thecis leviter longioribus per rimas elongatas dehiscentibus; stylo filiformi corollam subaequante, stigmate minute peltato. Santiago-Zamora: Eastern slope and crest of main Cordillera Cutuct, 5,200 ft. elev., Nov. 25=Dec. 2, 1944, H. Jorgensen CuJ]-38 (TYPE US 1,988,919; dupl. NY) (shrub 3 m.; flowers red). The generic position of this species presents difficulties not unfamiliar in the Andean Vacciniaceae. Although it is clearly of the general affinity of Thibaudia, it is perhaps best referred, because of its winged corolla, to the recently described genus Calopteryx (Jour. Arnold Arb. 27: 100. 1946), based on a single species from low elevation near the coast in El Valle, Colombia. From the only known species, C. insignis, C. sessiliflora differs in its leaf-blades with more highly concurrent secondaries, its greatly reduced l- or 2-flowered inflorescences sub- tended by conspicuous lanceolate bracts, its subsessile flowers, its much shorter and more narrowly winged corolla, and its anthers with comparatively short tubules. The discovery of this entity certainly weakens the characters which separate Calopteryx from Thibaudia, but the former concept may perhaps be maintained for the present. Thibaudia lateriflora A. C. Smith, sp. nov. Frutex scandens praeter flores ubique glaber, ramulis gracilibus teretibus inferne radicantibus; petiolis gracilibus rugulosis subteretibus 6-8 mm. longis crassis (2=3 mm. diametro); laminis subpapyraceis in sicco fusco-viridibus oblongo- ellipticis, 17-26 cm. longis, 7-10 cm. latis, basi obtusis, apice in acuminem gracilem ad 25 mm. longum subito caudato-attenuatis, margine inconspicue re- curvatis, subtus obscure et sparse glandulososstrigillosis, plerumque 7-nerviis, Mervis secundariis adscendentibus cum costa ad 3.5 cm. concurrentibus ut costa supra leviter impressis subtus prominentibus, nervis extimis inconspicuis, rete venularum .intricato supra subimmerso subtus prominulo; inflorescentiis in glome- tulos ramulis defoliatis enatos aggregatis, bracteis numerosis obscuris suffultis, thachi gracili minuta haud 2-3 mm. longa ut videtur plerumque uniflora, bracteis floriferis oblongis obtusis 1-2 mm. longis; pedicellis gracilibus 6-8 mm. longis ut calyce obscure puberulis basim versus bibracteolatis, bracteolis oblongis obtusis 1=1.5 mm. longis; calyce turbinato cum pedicello continuo sub anthesi 4~5.5 mm. longo et apice circiter 3 mm. diametro, obscure albido-puberulo etiam inconspicue glanduloso-strigilloso, tubo alis carnosis haud 0.3 mm. latis inconspicue alato, limbo erecto 1-1.5 mm. longo, lobis 5 inconspicuis ovato-denticulatis haud 0.5 mm. longis, sinibus obtusis; corolla tenuiter carnosa 5-angulata (angulis sub- alatis alis haud 0.2 mm. latis) sub anthesi 19-22 mm. longa circiter 4.5 mm. diametro obscure puberula mox glabrata, lobis 5 oblongis circiter 1.5 mm. longis; stamini- bus 10 corollam subaequantibus, filamentis ligulatis subcohaerentibus circiter 3 ad Y 76 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 mm. longis margine superne obscure puberulis,’antheris 18~19 mm. longis, thecis levibus 10-11 mm. longis, tubulis 8-9 mm. longis per rimas elongatas dehiscenti- bus; stylo filiformi leviter exserto, stigmate minute peltato. Santiago-Zamora: Cordillera Cutucu, ridge ascending into central Cutucd, 2,600 ft. elev., Nov. 17=Dec. 5, 1944, Camp E-1141 (coll. F. Prieto) (TYPE NY; dupl. US) (climbing; leaves dark green above, paler and shining beneath; calyx green-ribbed; corolla ribbed, crimson toward base, bright green distally). The fact that the calyx is continuous with the pedicel suggests that this species is an ally of the widespread T. floribunda H. B. K., from which it differs in such obvious respects as its broader leaves, greatly reduced and glomerulate inflorescences, narrowly winged calyx and corolla, and longer corolla and stamens. That the corolla is winged (although very obscurely so) indicates an affinity of the new species with Anthopterus Hook., where, however, the wings are obvious and manifestly veined. From another allied genus, Calopteryx A. C. Smith, T. lateriflora is excluded by its continuous calyx and also by its very narrow corolla wings. The discovery of such new species as this and the above-described Calop- teryx sessiliflora, however, suggests that the winged corolla cannot be safely utilized as a character of generic value in the family. Thibaudia clivalis A. C. Smith, sp. nov. Frutex epiphyticus ubique praeter antheras glaber, ramulis gracilibus junioribus obtuse angulatis vetustioribus teretibus; foliis subsessilibus, petiolis 1.5-3 mm. longis, laminis coriaceis subbullatis in sicco fusco-brunneis, oblongis vel obovato- oblongis 7.5=12.5 cm. longis, 4-8 cm. latis, basi manifeste cordatis et subauricu- latis, apice rotundatis, margine recurvatis et sinuato-crenatis, costa valida supra sulcata vel basim versus elevata subtus prominente, nervis lateralibus utrinsecus 3-5 supra impressis subtus valde elevatis, eis basim versus validioribus curvato- adscendentibus, rete venularum conspicuo utringue obtuse prominulo; inflores- centia completa non visa; pedicellis teretibus rugulosis ante anthesin circiter 10 mm. longis basi bibracteolatis superne incrassatis, articulatione conspicuo, bracteolis pentagonis circiter 2.5 mm. longis latisque subacutis; calyce turbinato 6-6.5 mm. longo apice 45 mm. diametro, tubo ruguloso obtuse angulato circiter 2.5 mm. longo, limbo suberecto carnoso quam tubo longiore 5-lobato, lobis late deltoideo-ovatis circiter 1 x 2 mm. apiculatis, sinibus obtusis; corolla (videtur paullo ante anthesin) urceolato-cylindrica carnosa 8-9 mm. longa basim versus circiter 4 mm. diametro superne angustata, lobis 5 deltoideis circiter 1 mm. longis acutis; staminibus 10, filamentis liberis ligulatis minutis ad 1 mm. longis margine superge pilosis, connectivis alternatis margine copiose albido-villosis (pilis circiter 0.5 mm. longis), alteris obscure puberulis, antheris circiter 6 mm. longis, thecis basi obtusis, tubulis quam thecis paullo brevioribus per rimas ovales magnas dehiscentibus; stylo tereti corollam subaequante truncato. Santiago~Zamora: Cordillera Cutucu, east-trending slope from top of ridge down toward the Itzintza, 4,800=5,800 ft. elev., Nov. 17-Dec. 5, 1944, Camp E-1384 (TYPE NY) (epiphyte; leaves dark green above, paler beneath, the veins red; young flowers pink). : Although the cited specimen is not entirely satisfactory, lacking attached or complete inflorescences, it-obviously represents an undescribed species of Thi- baudia. In its apparently rigidly carnose corolla, free filaments, erect elongate calyx-limb, and leaf-texture it suggests the Peruvian T. engleriana Hoer. From that species, however, T. clivalis differs in having its branchlets less sharply angled and its leaves subsessile, larger, and more deeply cordate at base. The flowers of the new species are not entirely mature, but they are comparatively 1952} PLANTS COLLECTED IN ECUADOR rar | small, with much shorter calyx-lobes, corollas, and stamens. Another species of this alliance, T. cardiophylla Sleumer, also of Peru, has the leaves smaller than those of the new species and the pedicels and comparatively large calyx densely pubescent. Thibaudia martiniana A, C. Smith, sp. nov. Frutex ubique filamentis exceptis glaber, ramis elongatis, ramulis robustis fuscis obtuse angulatis; petiolis crassis (3-5 mm. diametro) 13=20 mm. longis manifeste angulatis; laminis coriaceis in sicco fusco-olivaceis ovato-ellipticis, 20-30 cm. longis, 10-17 cm. latis, basi obtusis et in petiolum decurrentibus, apice ut videtur acutis vel breviter cuspidatis, margine integris et leviter re- curvatis, pinnatinerviis, costa supra paullo subtus valde prominente, nervis secundariis utrinsecus plerumque 3 curvatis supra conspicue impressis subtus prominentibus, rete venularum supra subplano vel prominulo vel leviter impresso subtus conspicuo; inflorescentiis axillaribus breviter racemosis circiter 10-floris sed floribus saepe caducis, rhachi crassa superne angulata circiter 1.5 cm. longa, bracteis caducis; pedicellis rugulosis sub anthesi 20—23 mm. longis superne valde incrassatis basim versus bibracteolatis, bracteolis ovato-deltoideis obtusis cir- citer 2 mm. longis, articulatione conspicuo; calyce coriaceo ruguloso cupuliformi- cylindrico sub anthesi haud apophysato 8-10 mm. longo apice 6-10 mm. diametro, tubo circiter 4mm. longo, limbo suberecto quam tubo longiore inconspicue 5-dentato, dentibus minute apiculatis haud 0.5 mm. longis, sinibus complanatis; disco carnoso pulvinato; corolla crasso-carnosa cylindrica sub anthesi circiter 25 mm. longa et basim versus 7 mm. diametro, superne gradatim angustata, lobis 5 deltoideis obtusis circiter 1.5 x 2.5 mm.; staminibus 10 corollam subaequantibus, filamentis liberis fuscis ligulatis circiter 5 mm. longis intus superne puberulis, antheris circiter 23 mm. longis crassis basi obtusis, thecis in tubulos 7-8 mm. longos in- ferne lateraliter connatos rimis elongatis dehiscentes gradatim transeuntibus; stylo crasso sub anthesi leviter exserto, stigmate obscure papilloso. Pichincha: Along the road from Quito to Sto. Domingo de los Colorados (western slope of the cordillera), 8,500-9,500 ft. elev., Jan. 15, 1945, Camp E-1717 (TYPE US 1,989,049; dupl. NY) (on banks in soil; branches arching to 4 m.; leaves dark green above, pale beneath, dull on both surfaces; pedicels red, the bracteoles white, the ‘‘joint’’ green; calyx red; corolla pure white). At the suggestion of the collector, this new species is named for Dr. William E. Martin, of the University of California, in whose company the type specimen was obtained. It is closely related only to T. pachypoda A. C. Smith, known from low elevations near the coast of El Valle, Colombia, but it differs in its more robust foliage, the leaf-blades being much larger and with more prominent second- ary nerves. The new species also has shorter pedicels, a slightly shorter corolla, and anthers with short tubules which are laterally adnate nearly to the apex. Another collection which may be mentioned here is Camp E-626 (NY only), from El Oro, in Moro-Moro region about 21 miles west of Portovelo, 3,400-4,200 ft. elev. (coarse shrub with branches to 3 m. long, in dense rain-forest). As compared with the type of T. martiniana, no. 626, which is past anthesis, has shorter petioles, pedicels up to 33 mm. long, and a shorter calyx-limb with more obvious lobes. Without corroborating evidence from the corolla and stamens, I hesitate to expand the concept of the new species to include this specimen, although it does not suggest any other known species. Thibaudia parvifolia (Benth.) Hoer. Bot. Jahrb. 42: 275. 1909. Azuay: ‘‘Oriente’’ Border, crest of Eastern Cordillera, between Ofia and the Rio Yacuambi, 10,000-11,200 ft. elev., F. Prieto P-305 (shrub 2 m.; leaves deep ‘ 78 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 1 green above, bright green beneath, nitid on both ‘surfaces; pedicels stout, green to red-tinged; hypanthium green to dull red; corolla dull reddish to deep crimson, pale distally; hypanthium and corolla with scattered gland-hairs), The species has apparently not otherwise been reported from Ecuador, but the cited specimen agrees excellently with Lehmann 2143 (US), from the type locality in Cauca, Colombia, and with several specimens recently obtained by Dr. Cuatrecasas in the Departments of Cauca and El Valle. Thibaudia jorgensenii A. C. Smith, sp. nov. Frutex parvus, ramulis gracilibus, junioribus obtuse angulatis pilis albidis circiter 0.5 mm. longis indutis, demum teretibus glabrescentibus; stipulis intra- petiolaribus subulatis circiter 2 mm. longis parce pilosis; petiolis subteretibus rugulosis 2-4 mm. longis ut ramulis mox glabratis; laminis coriaceis in sicco fuscis ovatis, 1.5-3 cm. longis, 0.8-1.5 cm. latis, basi leviter sed manifeste cordatis, apice obtusis, margine incrassatis et inconspicue crenulatis, supra copiose minute albido-punctatis, subtus parce et inconspicue glanduloso-strigillosis, e basi ob- scure 3-nerviis, Costa supra subplana subtus paullo elevata, nervis aliis basalibus et 1 vel 2 paribus e costa orientibus obscuris et immersis; inflorescentia apices ramulorum versus axillari subfasciculata ut videtur pluriflora bracteis imbricatis papyraceis pluribus suborbicularibus obscure piloso-marginatis maximis ad 7 mm. longis basi circumdata; pedicellis subteretibus 1-2 mm. longis pilis albidis 0.3-0.5 mm. longis copiose patenti-pilosis, basi bibracteolatis, bracteolis lineari-oblongis circiter 3 mm. longis margine ciliatis etiam parce glanduloso-pilosis caducis, articulatione manifesto; calyce campanulato sub anthesi circiter 9 mm. longo et apice diametro, tubo parvo haud 2 mm. longo ut pedicello copiose piloso, limbo erecto-patente papyraceo utrinque glabro vel extus inferne parce piloso profunde 5-lobato, lobis ovatis circiter 5 mm. longis (post anthesin ad 8 mm. accrescentibus) 2.5-4 mm. latis imbricatis manifeste nervatis acutis pilos glandulosos circiter 0.5 mm. longos margine gerentibus; corolla tenuiter carnosa cylindrica circiter 10 mm. longa et 4 mm. diametro utroque paullo angustata, extus ut calycis tubo copiose patenti-pilosa, lobis 5 deltoideis subacutis circiter 1 mm. longis; stamini- bus 10 circiter 7 mm. longis, filamentis liberis ligulatis 2.5-3 mm. longis superne anguStatis et margine pilosis, antheris 4.5-5 mm. longis, thecis tubulos longitu- dine subaequantibus basi inflexis et mucronulatis, rimis ovalibus circiter 1 mm. longis; stylo gracili corollam subaequante, stigmate minute peltato. Loja: Hda. Anganuma, at headwaters of Rio Cachiyacu, on west slopes of Cordillera Condor, about 46 km. south of Loja, 9,400 ft. elev., July 13-16, 1944, H. Jorgensen & F. Prieto ]P-47 (TYPE NY) (shrub, in sotobosque; leaves shining and deep green above, paler beneath; calyx deep pink to red; corolla white). The available material of the new species is not entirely satisfactory, but one good mature flower has been dissected. Because of its free filaments, the species would be sought among those numbered 10 to 18 in my key of 1932 (Contr. U. S. Nat. Herb. 28: 411), but its relationship is certainly with T. anomala A. C. Sm., based on an André collection without detailed locality. From this, T. jorgensenii differs in its leaf-blades with subcordate bases and its subsessile flowers with free filaments; T. anomala has the indument of the flowers much denser, covering even the calyx-limb, which in the new species is essentially glabrous. Cavendishia striata A. C. Smith, Jour. Arnold Arb. 27: 104. 1946, Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, between Bafios and Mera, 3,500-5,000 ft. elev., Camp E-2391 (NY only) (spreading shrub, terres- trial, up to 1 m. high; leaves pale green; bracts pinkish; flower-buds white). Santiago-Zamora: Cordillera Cutuci, on banks of Rio Itzintza, 3,500 ft. elev., 1952] PLANTS COLLECTED IN ECUADOR 79 Camp E-1207 (NY only) (short-branched epiphyte; bracts flushed with pink; hypan- thium greenish, the calyx-lobes white; corolla deep purple toward base, white above). This recently described species, occurring, like many others of the family, over a wide altitudinal belt in Pacific Colombia, has already been recorded from the Province of Pichincha but not elsewhere in Ecuador. In Colombia it occurs from near sea-level up to 2,000 m. The cited specimens both have very young in- florescences but present no important points of difference from Colombian material; under very high magnification the young bracteoles, calyces, and corollas are seen to be copiously glandular with minute spherical sessile glands. Cavendishia pseudospicata Sleumer, Bot. Jahrb. 71: 406. 1941. | Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, Shell Mera (east of Mera), about 3,500 ft. elev., Camp E-1702 (on steep bank, the branches drooping to 8 ft.; also seen as an epiphyte; ripe fruit purple-black, insipid). The cited specimen, in fruit, agrees excellently, in general, with Sleumer’s species, collected in the same region. However, the indument described by Sleumer appears to be fugacious, if my identification is correct. The following differences of no. 1702 from the original description should be noted: leaf-blades slightly larger (up to 10 x 3.5 cm.); inflorescence (except corolla, not seen) essentially glabrous in fruit except for a few scattered appressed glandular hairs; rachis slightly longer (to 11 cm. long). Cavendishia bracteata (R. & P.) Hoer. Bot. Jahrb. 42: 280. 1909. Thibaudia bracteata R. & P. Fl. Peruv. Chil. 4: pl. 388, 1802, ex J. St.-Hil. Expos. Fam. Nat. 1: 363. 1805. Proclesia hartwegiana Kl. Linnaea 24: 35, 1851, Cavendishia hartwegiana Hoer. Bot. Jahrb. 42: 489. 1909. Carchi: Cafion between San Gabriel and Bolivar, Camp E-375. Tungurahua: Along Rio Pastaza just west of Bafios, Camp E-2367, Cafiar: North rim of the valley of the Rio de Cafiar, between Suscal and Chontamarca, Camp E-2890 (coll. M. Giler). Valley of Rio de Cafiar at Abadel, below town of Galleturo, Prieto CP-33, Azuay: Numerous localities, Camp E-411, E-553, E-1984, E-2178, E-3936, E-4495, E-4892, E-5013, Loja: Cerro Villanaco (about 7 km. west of the city of Loja), Camp E-186-E-196 incl., E-682. Nudo de Cajanuma, 7 km. south of Loja, Camp E-114, E-115, Crest of the Cordillera de Zamora, east of Loja, Camp E-93, E-94, E-103, Napo-Pastaza: Valley of the Rio Pastaza and adjacent uplands, near El Topo, along trail to La Gloria, Camp E-2400, Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza, near mouth of Rio Patos, Camp E-754., The extensive suite of specimens cited above (without detailed notes on precise locality, habit, habitat, and color, since such notes are too abundant for inclusion here) demonstrates considerable variation, and yet I hardly see how it can be re- ferred to more than a single species. The material is from shrubs up to 5 m. high, taken at elevations from 4,000 to 10,000 ft., and the corolla color is uniformly noted as red to crimson below, yellow or greenish yellow at apex. The most strik- | ing variations are seen in the indument of corolla and calyx, and in the density of glands on the calyx, pedicels, bracts, etc. The corolla, in particular, varies from copiously white-pilose with short spreading hairs to entirely glabrous. Of particular value is Dr. Camp’s series numbered E-186-E-196 inclusive, taken from plants in the same colony, demonstrating the instability of these: characters. In my treatment of 1932 (Contr. U. S. Nat. Herb. 28: 489 et seq.) I expressed uncertainty as to the biological validity of several species of this alliance, and a hartwegiana in that work was keyed with both the pubescent- and the glabrous- 80 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [ Vol, 8, No. 1 flowered species. It now seems quite impossible to maintain C. hartwegiana as distinct from C. bracteata. The accumulation of herbarium material in the past twenty years has furthermore served to weaken the supposed distinctions between C. bracteata and several other of its allies. One may question the advisability of maintaining such species as C. beckmanniana Hoer. (1909), C. scabriuscula (H. B. K.) Hoer. (based on Thibaudia scabriuscula H. B. K., 1818), and C. micom- oides A. C. Smith (based on Thibaudia melastomoides H. B. K., 1818). The com- plex of which these entities are a part (whether as species or taxa of lesser rank) extends from Colombia to Bolivia. It should also be considered whether the C. bracteata complex can be specifi- cally kept apart from C. strobilifera (H. B. K.) Hoer. (based on Thibaudia strobili- fera H. B. K., 1818), to which I have already reduced many comparatively recent specific concepts (i. e. to C. acuminata, op. cit. 503-505). Certainly I should now refer to C. bracteata several of the Ecuadorian collections which in 1932 I cited as C. acuminata. The problems of relationships in this group of Cavendishia can probably not be solved without analysis in the field. In referring Dr. Camp’s material to C. bracteata I make use of the oldest specific epithet for the complex. Cavendishia strobilifera (H. B. K.) Hoer. Bot. Jahrb. 42: 279, 1909, Pichincha: Western slope of the cordillera, along the road from Quito to Sto. Domingo de los Colorados, 7,000-10,000 ft. elev., Camp E-1720 (abundant shrub 0.3-2.5 m.; bracts crimson; hypanthium red; corolla crimson except for pale yellow apex; ripe fruit purpleblack, shining). As implied above in my discussion of C. bracteata (R. & P.) Hoer., the differ- ences between that and C. strobilifera are not very convincing. Since the character of corolla-pubescence is seen to be of little use, only the somewhat larger leaves with definitely long-acuminate apices serve to keep the present species apart. Cavendishia capitata (Benth.) Hoer. Bot. Jahrb. 42: 279, 1909, Loja: ‘‘Oriente’’ Border, crest of the Cordillera de Zamora, east of Loja, ca. 10,000 ft. elev., Camp E-102 (growing in soil; spreading shrub to 2 m. high;:bracts red; corolla white). Santiago-Zamora: Valley of the Rio Zamora, east of Loja, ridge across river from village of Zamora, 6,500 ft. elev., Camp E-42, Eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), between Tres Ranchos and Chontal, 2,700-5,700 ft. elev., Camp E-1557 (great mounds of canes on bank, probably starting as windfall; leaves shining on both surfaces; bracts crimson; corolla white). The cited specimens agree excellently with the type, from the Province of Loja. The collections discussed here represent the only additional material of the species known to me. The species is characterized by its essentially glabrous habit, comparatively large leaves and flowers, short pedicels with conspicuous bracteoles, glandular calyx-limb, and comparatively long glandular-margined calyx-lobes. The three cited collections permit some amplification of my earlier description (Contr. U. S. Nat. Herb. 28: 507. 1932), as follows: Leaf-blades (9=)12-19 cm. long, (3.5-)4.5-8.5 cm. broad, 7- or 9-nerved from near base but the outer 2 or 4 nerves often very inconspicuous; pedicels 2=5 mm. long, sometimes with a few scattered minute spherical glands, bibracteolate toward base, the bracteoles ‘linear-oblong, obtuse, 4.5-6 x 1=1.5 mm., glandular- margined and sometimes glandular on both surfaces distally; calyx at anthesis up to 13 mm. long and 9 mm. in diameter at apex, the tube angled, up to 7 mm. long, the limb 3-6 mm. long, bearing superficial spherical glands without, the lobes 2.5-5 x 3=4 mm., glandular-margined, the sinuses rounded or obtuse; corolla 1952] . PLANTS COLLECTED IN ECUADOR 81 22-30 mm. long (as previously described), 4-7 mm. in diameter; filaments alter- nately 3.5-5 mm. and 6-8.5 mm. long, the anthers alternately 18-19 mm. and 15-17 mm. long, with thecae 5-7 mm. long. The following three specimens should also be considered as representing C. capitata, although they are not strictly typical: Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), between Hda. Chontal and Sta. Elena, 3,400-4,600 ft. elev., Camp E-804 (NY only) (arching shrub, often seen epiphytic; bracts red; corolla pale pink, distally white). Cordillera Cutuci, ridge just south and west of Rio Itzintza, 5,900 ft. elev., Camp E-1347 (NY only) (on soil in moss; fruit black at maturity, insipid). Eastern slope and crest of main Cordillera Cutuct, 5,500 ft. elev., Jorgensen Cu] -43 (shrub 3 m.; corolla red). These specimens have the inflorescence precisely as in typical C. capitata, except that the pedicellary bracteoles are inclined to be slightly smaller (2.5-3.5 mm. long). The leaf-blades are comparatively narrow and lanceolate-oblong, 8=13 cm. long and 3-4 cm. broad, being sometimes only 5-nerved. These differences are so inconsequential that a reasonable species-concept for C. capitata may include all six specimens cited above. Cavendishia campii A. C. Smith, sp. nov. Frutex ad 4 m. altus, ramulis fuscis subteretibus apicem versus pilis albidis 0.2-0.5 mm. longis indutis demum glabratis; petiolis rugulosis 7=11 mm. longis ut ramulis pilosis; laminis coriaceis in sicco fusco-olivaceis oblongo-ellipticis, 8-17 cm. longis, 2.5-6 cm. latis, basi anguste rotundatis vel obtusis, in acuminem gracilem subacutum 1-2 cm. longum angustatis, margine recurvatis, supra glabris vel basim versus obscure pilosis, subtus pilis castaneis circiter 0.2 mm. longis e basi incrassato adpressis copiose strigillosis ac etiam secus nervos albido- hispidulis, 5(vel obscure 7-}nerviis, nervis secundariis basi (raro ad 1 cm. con- currentibus) orientibus ut costa supra impressis subtus prominentibus, rete venularum supra plano subtus obscuro; inflorescentia axillari compacta plerumque 8=12-flora bracteis numerosis papyraceis glabris oblongis ad 2.5 cm. longis basi circumdata, bracteis extimis (minimis) dorso parce pilosis, rhachi crassa 0.5<2 cm. longa mox glabra; pedicellis rugulosis sub anthesi 4-9 mm. longis parce pilosis glabrescen- tibus, basim versus bibracteolatis, bracteolis lanceolato-oblongis subacutis cir- citer 4 mm. longis glabris vel obscure glanduloso-marginatis; calyce sub anthesi 6-9 mm. longo et apice 5-6 mm. diametro extus albido-piloso et interdum minute glanduloso, tubo oblongo obtuse angulato 4-5 mm. longo, limbo erecto 2=5 mm. longo 5-lobato, lobis deltoideis. 1.5=3 x 2-3 mm. subacutis interdum glanduloso- marginatis, sinibus acutis vel obtusis; corolla tenuiter camosa cylindrica sub anthesi 15-21 mm. longa et 4-5.5 mm. diametro praeter basim et apicem pilis 0.3-0.4 mm. longis patentibus ornata, lobis oblongis obtusis circiter 1 mm. longis; staminibus alternatim 13-14 mm. et 15-16 mm. longis, filamentis liberis gracilibus alternatim 2.5=3 mm. et 5=5.5 mm. longis superne pilosis, antheris alternatim circiter 12 mm. et 11 mm. longis, thecis 3-5 mm. longis; stylo corollam subaequante. Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), between Hda. Chontal and Sta. Elena, 3,400-4,600 ft. elev., Nov. 1, 1944, Camp E-803 (TYPE US 1,989,003; dupl. NY) (shrub 3 m.; bracts deep pink; corolla red), Camp E-785 (NY only) (arching shrub 4 m.; bracts deep pink). Azuay: The eastem Cordillera, 1-8 km. north of the village of Sevilla de Oro, 8,000-9,000 ft. elev., Camp E-4350 (coarse shrub, with branches to 4 m.; leaves deep green and nitid above, paler beneath; bracts pink; base of corolla pale pink, central portion deep pink to pale rose, 82 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vor 8, No. 1 apex white). Quebradas leading into the Rio Collay, 3=<8 km. north of Sevilla de Oro, 7,000-8,300 ft. elev., Camp E-5007 (sprawling shrub 3 m.; leaves deep green above, pale beneath, subnitid on both surfaces; bracts pale pink; corolla basally pink, apically white). In my key to the genus (Contr. U. S. Nat. Herb. 28: 463-467, 1932) this species would be sought in the vicinity of C. pubescens (H. B. K.) Hemsl., from which it differs in its less copious indument, generally smaller leaves, more compact inflorescence, and smaller flowers with less conspicuous calyx-lobes. However, the new species is probably more closely allied to C. capitata (Benth.) Hoer., discussed above. From the typical form of C. capitata, C. campii differs in the pubescence of its vegetative parts and flowers, its prevailingly smaller leaves, longer pedicels with smaller bracteoles, eglandular (or sparsely glandular) calyx, and distinctly shorter corolla and stamens. Variability in pubescence is marked in some species of Cavendishia (see discussion of C. bracteata, above), but the combination of characters marking C. campii seems reasonably adequate, although admittedly specific lines in this group of Cavendishia are somewhat arbitrary and in need of field analysis. Another species of this alliance, the Peruvian C. ulez Hoer., has its leaf-blades distinctly 7-nerved and with comparatively highly con- current nerves, and lacks the characteristic white pubescence of C. campit. Cavendishia zamorensis A. C. Smith, sp. nov. Frutex epiphyticus, ramulis crassis glabris in sicco pallidis striatis; petiolis subteretibus rugulosis glabris 5-10 mm. longis; laminis subcoriaceis siccitate metallico-olivaceis ellipticis, 1l+15 cm. longis, 5-9 cm. latis, basi rotundatis, in acuminem subacutum ad 15 mm. longum subito angustatis, margine recurvatis, supra glabris, subtus minute glanduloso-strigillosis, 7(vel obscure 9-)-nerviis, nervis (intimis interdum ad 2 cm. cum costa concurrentibus) et costa supra im- pressis (vel basim versus elevatis) subtus prominentibus, rete venularum utrinque haud prominulo; inflorescentia apices ramulorum versus axillari subcapitata, ubique praeter filamentas glabra, bracteis papyraceis subofbicularibus rotunda- tis margine scariosis maximis ad 2 cm. diametro basi circumdata, rhachr brevi, floribus numerosis congestis; pedicellis incrassatis sub anthesi 2=3 mm. longis parce glandulosis basim versus bibracteolatis, bracteolis submembranaceis oblongo- ellipticis 5-6 mm. longis circiter 4 mm. latis apice rotundatis utrinque parce glanduloso-strigillosis tubum calycis involventibus; calyce sub anthesi circiter 10 mm. longo et apice diametro, tubo obtuse angulato circiter 3 mm. longo, limbo erecto-patente quam tubo longiore supeme parce glanduloso-strigilloso profunde S-lobato, lobis elliptico-oblongis basi imbricatis 6=7 mm. longis 4-5 mm. latis apice rotundatis margine scariosis et pauci-glandulosis; corolla carnosa urceolato- cylindrica sub anthesi 13-14 mm. longa et circiter 7 mm. diametro, superne angustata, lobis deltoideis obtusis circiter 1 mm. longis; staminibus circiter 11 mm. longis, filamentis ligulatis alternatim circiter 2 mm. et 3.5 mm. longis superne intus obscure pilosis, antheris alternatim circiter 11 mm. et 10 mm. longis, thecis 4~5 mm. longis; stylo tereti corollam subaequante, stigmate minuto. Santiago-Zamora: Valley of the Rio Zamora, east of Loja, near Zamora, about 3,000 ft. elev., June 28=July 1, 1944, Camp E-2 (TYPE US 1,988,931; dupl. NY) (epiphyte on trees over river; bracts deep pink to crimson; corolla white). In its imbricate calyx-lobes, C. zamorensis suggests a relationship with such Colombian species as the recently described C. tenella A. C. Smith (Jour. Arnold Arb. 27: 106. 1946), from which it differs in its short-petiolate leaves with rounded bases, compact and comparatively few-flowered inflorescences, large pedicellary bracteoles, and slightly larger flowers. A specimen which may best be referred to 1952] PLANTS COLLECTED IN ECUADOR 83 the new species, but the variations of which are not included in the above de- scription, is Camp E-29 (NY only) (same locality as type, ridge across the river from the village of Zamora, 6,500 ft. elev.; epiphytic shrub with branches to 3 m.; bracts deep pink; flowers pale pink). From the type this specimen differs in having the pedicellary bracteoles only 1.5 mm. broad and consequently not clasping the calyx, the calyx-lobes only 2.5-3 mm. broad, very narrowly imbricate, and more copiously glandular at margin, the corolla 19-22 mm. long, the stamens about 15 mm. long, with filaments alternately about 2.5 mm. and 5 mm. long and stamens about 13 mm. and 11 mm. long respectively. This variant seems to point toward a relationship of C. zamorensis with C. capitata, which occurs more commonly in the area, as noted above. The type of the new species, at least, seems to represent an entity worthy of specific rank, differing from typical C. capitata in its broad bracteoles and calyx-lobes and its short corolla and stamens. Cavendishia orthosepala A. C. Smith, sp. nov. Frutex epiphyticus, ramulis subteretibus pilis albidis circiter 0.5 mm. longis patentibus indutis demum glabratis; petiolis validis teretibus 7-13 mm. longis ut ramulis pilosis; laminis subcoriaceis plus minusve bullatis in sicco fusco-metallicis oblongo-ellipticis, (11-)15=26 cm. longis, (3=)4-9 cm. latis, basi rotundatis vel late obtusis, in acuminem subacutum 10=15 mm. longum subito angustatis, margine valde recurvatis, supra primo pilosis, subtus ut ramulis copiose albido-pilosis ac etiam pilis castaneis glandulosis circiter 0.2 mm. longis a basi incrassato ad- pressis strigillosis, plerumque 7-nerviis, nervis (intimis ad 3 cm. cum coSta con- currentibus) et costa supra conspicue impressis subtus prominentibus, venulis utrinque subprominulis; inflorescentia subterminali breviter racemosa multiflora, bracteis -_papyraceis obovato-oblongis rotundatis margine scariosis maximis cir- citer 25 mm. longis basi circumdata, bracteis extimis (minimis) dorso pilosis ceteris glabris, rhachi post anthesin ad 3 cm. longa glabra; pedicellis incrassatis teretibus obscure..glanduloso-strigillosis 3-5 mm. longis basi conspicue bibracteo- latis, bracteolis papyraceis obovato-ellipticis 12-16 mm. longis 3-6.5 mm. latis apice rotundatis vel emarginatis et parce glandulosis margine scariosis dorso glanduloso-strigillosis tubum calycis involventibus; calyce post anthesin 15-17 mm. longo extus pilis albidis 0.4-0.7 mm. longis copiose induto, tubo cupuliformi circiter 5 mm. longo et diametro, limbo erecto fere ad basim profunde 5-lobato, lobis subcoriaceis anguste oblongis 10=12 mm. longis 2=3 mm. latis basi anguste imbricatis apice subacutis intus glabris glandulososstrigillosis; corolla filamenti- sque non visis, antheris circiter 10 mm. longis, tubulis thecas longitudine sub- aequantibus; stylo filiformi circiter 16 mm. longo, stigmate minuto. Santiago-Zamora: Cordillera Cutuct, on banks of Rio Itzintza, 3,500 ft. elev., Nov. 17=Dec. 5, 1944, Camp E-1201 (TYPE US 1,989,011; dupl. NY) (epiphyte; leaves deep green and shining above, paler and dull beneath; inflorescence up to 24-flowered; outer bracts deep red, the inner bracts pink to red; fruit white). Same locality, ridge between Rios Itzintza and Chupiasa, 4,000-4,500 ft. elev., Camp E-1278 (vine-like epiphyte, in tree thrown by storm; leaves deep green above, paler beneath, with markedly translucent veins; bracts crimson; hypanthium turn-— ing cream-white, the calyx-lobes crimson). The remarkably developed pedicellary bracteoles and calyx-lobes of the new species are so unique in Cavendishia that comparisons with other species are superfluous. A tendency in this direction is noted in the above described C. zamorensis, which differs from C. orthosepala in the lesser development of these parts and in obvious characters of pubescence, but which may be its closest relative. The new species suggests C. pubescens (H. B. K.) Hemsl. in foliage, ‘ 84. MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol.*8, No. 1 but inflorescence differences are numerous and obvious. No corollas are available for the new species, but dimensions of the style and a few anthers (with no. 1278) indicate a corolla about 16 mm. long. Cavendishia sp. Santiago-Zamora: Cordillera Cutuct, ridge ascending into central Cutuci, 4,400-4,700 ft. elev., Camp E-1157 (NY only) (epiphyte, arching to 2 m.; calyx fluted fad ribbed; corolla pale, tipped with crimson). The cited plant appears to represent an undescribed species, but unfortunately the corollas described in the field notes are not now with the unicate specimen. Its relationship is with the species of Cavendishia with an elongate calyx-limb and callose-thickened lobes (see sp. 5~9 in my key in Contr. U. S. Nat. Herb. 28: 463. 1932), Orthaea secundiflora (Poepp. & Endl.) Kl. Linnaea 24: 24, 1851. Santiago-Zamora: Easter slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro, to Mendez), between Hda. Chontal and Sta. Elena, 3,400-4,600 ft. elev., Camp E-783 (coarse epiphyte, hanging lax to 6 m.; new leaves bright pink, conspicuous in the forest; corolla deep crimson, in bud pale pink, apically white); same locality, Camp E-799 (arching shrub 4 m.; corolla crimson below, apically white). These collections, the first of the species recorded from Ecuador, agree excellently with the original pea eee and plate (Thibaudia secundiflora Poepp. & Endl. Nov. Gen. & Sp. 1: 5. pl. 9. 1835). The type material of the species was obtained in the present Department of Huanuco, Peru. Orthaea sp. Santiago-Zamora: Eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), region of Tambo Pilas, near mouth of Rio Patos, 6,500-7,500 ft. elev., Camp E-757 (shrub 4 m.; leaves shining on both sides, somewhat paler beneath; corolla pale pink below, apically white). The cited specimen differs from O. secundiflora in its congested inflorescences (rachis 5=8 mm. long; pedicels about the same length), and its stamens with free filaments and slightly shorter anthers. It may be referable to O. abbreviata Drake, from southern Ecuador, but the inadequate description of that species (Jour. de Bot. 3: 75. 1889) implies that its leaves are somewhat broader (7 x 3 cm.) and its pedicels are 2 cm. long. Number 757 cannot be positively identified without comparison with the type of O. abbreviata. Satyria” panurensis (Benth.) Benth. & Hook. Gen. Pl. 2: 568, 1876, Santiago-Zamora: Low hills west of Rio Upano, along Rio Chupiangias, 2,500- 3,200 ft. elev., F. Prieto ChuP-24 (epiphytic, vinelike; young leaves red; corolla red, green at apex). Widespread in the Andean foothills from Peru to Venezuela and into British Guiana, but not, so far as I know, previously recorded from Ecuador. Satyria leucostoma Sleumer, Bot. Jahrb. 71: 407, 1941. Santiago-Zamora: Cordillera Cutuct, ridge ascending into central Cutuci, 4,400-4,700 ft. elev., Camp E-1146 (climbing, or high epiphyte; pedicels and hypanthium green; base of corolla crimson, the apex white; leaves dull beneath). Eastern slope of the cordillera, valley of the Rios Negro and Chupianza (on the trail from Sevilla de Oro to Mendez), between Tres Ranchos and Chontal, 2,700- 5,700 ft. elev., Camp E-1558 (epiphyte; leaves deep green above, pale beneath; pedicels and hypanthium green; base of corolla red, the apex white). 1952] PLANTS COLLECTED IN ECUADOR 85 Although type material of Sleumer’s species is not available, the cited speci- mens agree excellently with the original description, based on Schultze-Rhonhof 3010, also collected in the Oriente of Ecuador. Some of the leaf-blades of the Camp material, probably from older parts of the plant, attain dimensions of 30x 9 cm.; the corollas are 8.5=-9.5 mm. long, and the anthers (slightly larger than those of the type) are alternately 3.5-4.5 mm. and 4-5 mm. long. DEPARTMENT OF BOTANY, U. S. NATIONAL MUSEUM SMITHSONIAN INSTITUTION, WASHINGTON, D. C. Pips ey 7 a te 4 +h fs carat Pe o4a thr aa sYagiica? fon ae | gotten tig iat ge ade Abbniesan ty ae ye! It ott ‘ ” a? 4n Oe © Fer gy segte ia! “RS Ae wis veg gee) olaigels ered Sen ; 7 "ei “* Pe k3% U * ‘ ‘ = - Wa i : = ~_) ma on ul 4 a | ares > 4 * 2 _ : ‘ < ’ - i / Sg a . ‘ ' : = - he cd ; aes - ie oa _ =. 2 > 2 fe ook oa) a a ad is — x . al ~ ? s lie’ a ~ ies _-— = se i , a a —— s Z : = 5 ball ae 1s 3 C 2 3 = - x a ie . 4 ; 7 poe a , a. a. ae ch 2-75 =i ‘. ge aks x 7 aie, eI Mages aed are es ei a poe“, oll st palin ee te sah Tod 3d ae ~S 7 ices qa alee hs ee? Siedtk ih: a 2 x i ia a Mm: a eget rag : om Fi } ee a Ges ay Rants <8 is - hop ng? | hens PRS ret... p ae PUBLICATIONS OF THE NEW YORK BOTANICAL GARDEN Flora of Bermuda, by Nathaniel Lord Britton. 585 pages, 519 figures. 1918. $3.50. The Bahama Flora, by Nathaniel Lord Britton and Charles Frederick Millspaugh. 695 pages. 1920. $6.25. Flora of the Prairies and Plains of Central North America, by P. A. Rydberg. 969 pages, 601 figures. 1932. $6.00. Plants of the Vicinity of New York, by Henry Allan Gleason. 284 pages, illustrated. 1935. $2.00. An Illustrated Flora of the Northern United States, Canada and the British Posses- sions, by Nathaniel Lord Britton and Addison Brown. Including descriptions and illustra- tions of 4,666 species. Three volumes. Second edition, reprinted, 1936. $15.00. North American Cariceae, by Kenneth Kent Mackenzie. 539 figures of Carex and re- lated species by Harry C. Creutzburg, with a description of each. Two volumes. 1940. $17.50. Keys to the North American Species of Carex, by Kenneth Kent Mackenzie. From North American Flora, Vol. 18. 1941. $1.25. Vegetables and Fruits for the Home Garden. Four authoritative articles reprinted from the Journal of The New York Botanical Garden, edited by Carol H. Woodward. 21 pages, illustrat ® 1941. 15 cents. Food and Drug Plants of the North American Indian, by Marion A. and G. L. Wit- trock. Two articles, illustrated, from the Journal of The New York Botanical Garden. 1942. 15 cents. Catalog of Hardy Trees and Shrubs. A list of the woody plants in cultivation outdoors at The New York Botanical Garden in 1942. 127 pages, map and 20 illustrations. 1943. 75 cents. Succulent Plants of New and Old World Deserts, by E. J. Alexander. 350 species, 100 illustrated. 64 pages. Third edition, 1950. Bound in paper, 50 cents. The Flora of the Unicorn Tapestries, by E. J. Alexander and Carol H. Woodward. 28 pages, illustrated. Third edition, 1950. Bound in paper, 25 cents. Periodicals Contributions from The New York Botanical Garden. Technical papers written by students and members of the staff and reprinted from various journals. Fourteen volumes. 1899-1913. 25 cents each, $5.00 a volume. Memoirs of The New York Botanical Garden. Issued irregularly. 1900- Vols. 1-6, to me »bers of the Garden, $1.50 a volume; to others, $3.00. Vol. 7, to members, $2.50; to others, $5.00. Vol. 8, $10.00. Journal of The New York Botanical Garden. Notes, news, and non-technical articles on botany and horticulture, illustrated. (Replaced by The Garden Journal of The New York Botanical Garden.) 51 volumes. 1900-1950. North American Flora. Descriptions of the wild plants of North America, including Greenland, the West Indies, and Central America. Issued irregularly; 100 parts have ap- peared. 1905- Prices of parts on request. Mycologia. Issued bimonthly; devoted to fungi; official organ of the Mycological Society of America. Now in its forty-third volume. 1909 $1.50 a single copy, $7.00 a volume. __ Addisonia. Colored plates of flowering plants, accompanied by popular descriptions; eight plates in each number, thirty-two in each volume. Now in its twenty-second volume. 1916- Free to members of the Garden; to others $10.00 a volume. Brittonia. Issued irregularly; devoted to taxonomy and related subjects. Official organ of the American Society of Plant Taxonomists. Now in its seventh volume. 1931- To members of the Society, $5.00 a volume; to others, $7.50. The Garden Journal of The New York Botanical Garden. (Replaces the Journal of The New York Botanical Garden.) Issued bimonthly; contains non-technical articles on bot- oo and horticulture, illustrated. 1951- Free to members of the Garden; to others $2.00 a volume. A > > ‘eS N Sty —_ ? ¢ ‘! 1) si : ‘ * —< ’ £ > a fi ap ae a dene i= al te : AS “> ss > =i " - — > ee / se = mr i>. 4... . Pte ~~" + oy Aas og Dy mahal a t i ye Poss - % a vane - i oe Sa < \ a * - > 4 > ry Wy = a a . : ; = pars , ~~ “ eal Sar - La. eo" eas oy, ‘ mt MEMOIRS OF THE NEW YORK BOTANICAL GARDEN VoL. 8, No. 2 The Botany of the Guayana Highland Bassett Maguire, Richard S. Cowan, and John J. Wurdack, and Collaborators 87 Issued 27 April 1953 PRINTED BY THE SCIENCE I’REsS LANCASTER, Pa, The Memoirs of The New York Botanical Garden are issued at irregular intervals in parts of various sizes. Approximately 500 pages will complete a vol- ume. The subscription price of volume 8 is $10. Number 2 may be purchased separately for $2.00. Authors of papers may obtain separate copies of their contributions, printed at the same time as the issue, at cost price. For further information address the editor: H. W. RickETT The New York Botanical Garden New York 58, N. Y. Vol. 8 No. 2 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN April 27, 1953 THE BOTANY OF THE GUAYANA HIGHLAND A Report of the Kunhardt, the Phelps, and the New York Botanical Garden Venezuelan Expeditions BASSETT MAGUIRE, RICHARD S. COWAN AND JOHN J. WURDACK, AND COLLABORATORS INTRODUCTION Seldom do richness of natural history and delightful fantasy of fabulous legend and classic literature so combine to lure attention as they do in the ancient region of Guayana. Barely ten years after Columbus’ discovery of the New World, tales were told in maritime Europe of incredible men and practices in what was soon to be called Guayana; of headless savages, warewaging women, and frightful arrow- and dart- poison (curare). The myth of the Mar del Dorado somewhere in the remote interior where the Indians covered themselves with dust of gold gained wide currency in Europe. Raleigh a hundred years later led four expeditions into the mouth of the Orinoco in quest of empire and in search of the gold of Dorado. Another two hundred years had passed before von Humboldt had skirted the northern and western periphery of Guayana. And it was not until 1838-39 that Robert Schomburgk traversed the wild region from east to west by foot and canoe, giving meaning to the geography of Guayana, and finally quieting the stories of the gold of El Dorado. But Indian legend and mythology continued to flourish. Such stories against the backdrop of fantastic grandeur of landscape, in the light of newly rising con- sciousness of the existence of an extraordinary flora and fauna on the isolated mountain fastnesses, lent themselves to the beautiful writings of William H. Hud- son and the imaginative classic of Conan Doyle. Actual facts of physical history and diversity and extraordinary endemism among the biota of the isolated and lofty sandstone plateaus are found to be hardly less spectacular than precursor fable and fantasy. It was these attractions, largely revealed by the discoveries and writings of the Schomburgk brothers, that during the latter part of the nineteenth century drew European biologists to Mt. Roraima, the ‘*Lost World’’ of Doyle. It is still that inexhaustible treasure-house of natural history that continues to draw to the Guayana Highland naturalists from the Americas and Europe, particularly from Venezuela and the United States. Since 1925 continued exploration into still little known Guayana has been carried on chiefly by Venezuelan naturalists, notably the Phelps, father and son, distin- guished ornithologists; by Captain Felix Cardona, explorer-extraordinary for the Venezuelan Government; by English botanists in the Kaieteur region; and by the . cooperative efforts of the Servicio Botanico, Caracas, the American Museum of Natural History, the American Geographical Society, the British Guiana Forest De- partment, and the New York Botanical Garden. These reports detail the results of the continued exploration of the New York Botanical Garden and associated insti- tutions and agencies, and may be considered preliminary to a floristic treatment of the phytogeographic province, the Guayana Highland. There follows in this introduction a brief statement of the pertinent botanical history of the region. Any consideration of geology and geography, and any inter- pretation of phytogeography and ecology of the flora may more successfully be made at the conclusion of the continuing program of exploration and the comple- tion of the review of collected material. 87 . 88 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 History of Botanical Exploration. Active exploration of the Guayana Highland began with the extraordinary travels (mostly confined to British Guiana) of the famous brothers, Robert Herman and Richard Schomburgk. Robert Schomburgk’s remarkable journey of exploration in the Guayana High- land started on September 20, 1838, on his departure from the Brazilian garrison, Fort Sao Joaquim, at the junction of the Takutu and Parima where together they form the Rio Branco. Under a commission of the Geographical Society of London Schomburgk was to proceed westward overland, finally to reach the Orinoco River, and thus supplement with his own the observations of von Humboldt made along the Orinoco more than thirty-five years earlier. It was to be seven months and two days before Schomburgk again reached Fort Sao Joaquim. He had become the first botanist to visit the fabulous Mount Ro- raima. In his further westward progress he had crossed and recrossed the low Pacaraima Divide separating the drainages of the Orinoco to the north and the Amazon to the south; he visited the remote Cerro Marahuaca and its more westerly neighbor Cerro Duida on the Orinoco. And, after having passed through the Casi- quiare, linking the waterways of the Orinoco with the-Amazon, finally by way of the Rio Negro and Rio Branco he completed his ‘‘circular tour of 2200 miles’’!’ This certainly is one of the great explorations of all time. Again in 1842 Robert Schomburgk returned to Mt. Roraima, this time with a party including his younger brother Richard. This expedition set out on the morn- ing of September 10, 1842,’ from the granitic Kanuku Mountains in west-central British Guiana. At the Ireng River, only some 20 miles from Pirara, the party reached the Brazilian frontier, and from the junction of the Ireng with the Takutu, a few miles below, were wholly out of British Guiana until the return to that point before arriving again at Pirara on the following December 28th. The plant (and animal) specimens collected on this important exploration were, as a consequence, taken from either Brazilian or Venezuelan soil, except those that may have been obtained by Robert after he, independently, had crossed the present Venamo-Roraima boundary line from Venezuela into the British Guiana drainage of the Kamarang River. The title of Schomburgk’s journal, ‘*Travels in British Guiana,’’ does not indi- cate extra-British Guiana excursion. Locality on collection labels is indicated by the vague designations ‘‘British Guiana’’ or ‘‘Roraima’’; however, in the narrative (op. cit. pp. 118-277) itinerary and locality were plainly given. The Schomburgks had known fairly accurately where they had been since they were observers of much of the border disturbances of the time, and Robert himself had been a mem- ber of the Boundary Commission. Since the excursions of the Schomburgk brothers, Roraima has attracted numer- ous explorations. Burkill recounts the history of botanical exploration of the mountain up to 1900 in the important ‘*Report on two Botanical Collections Made by Senate F. V. McConnell and J. J. Quelch at Mount Roraima in British Gui- ana.’’* The following is taken largely from Burkill’s account. Karl Appun visited the mountain in 1864 (Burkill’s account p. 2) where baa reached the base of the cliff on the east and south side. Im Thurn and Perkins (p. 3) were the first to ascend the mountain to the summit, but were forced to re- turn to their base camp at 5400 feet after a distressingly short stay of three hours. *Schomburgk, R. H. Travels in Guiana and on the Orinoco, 1841. English translation by Walter E. Roth, Georgetown, 1931. *Schomburgk, Richard. Travels in British Guiana 2: 118. 1848. English translation by Walter E. Roth, Georgetown, 1923. . 3Brown, N. E, et al. Trans. Linn. Soc. II. 6: 1-107. pL 1-14. 1901. , a 1953] BOTANY OF THE GUAYANA HIGHLAND 89 McConnell and Quelch twice visited Roraima, in 1894 and again in 1898, both times reaching the summit, where a total of 12 days were spent making animal and plant collections. They had crossed the Ireng at Karona Falls into Brazil from British Guiana. Roraima was approached from the south up the valley of the Ara- bapo in Venezuela, and ascended by way of the Kukenam drainage in Venezuela along the southwest face of the mountain. Their extensive collections form the basis for the report by N. E. Brown which remains today the most important single contribution to the botany of Roraima. Unfortunately, although locality is given in the citation of specimens, by indirection from the title of the work one is led to consider that the collections were made in British Guiana, whereas in fact the great bulk was made in Venezuela and a lesser amount in Brazil. Gleason* has reported that ‘‘E. Ule spent several weeks on the slopes of the mountain [Roraima]l in December 1909 and January 1910, and made four ascents to its summit.”’ but that (Gleason p. 406) ‘tno summary has been made of the number of species collected by Ule.”’ Contemporary botanical study of Roraima began with the Lee Garnett Day— American Museum of Natural History Expedition led by G. H. H. Tate of the Amer - ican Museum. Roraima was approached by the way of the Amazon and Rio Branco, and ascended by the then well established trail up the southwestern face. Collec- tions on Roraima and its approaches were made from October 27, 1927, to January 9, 1928. Although the primary purpose was the study and collection of the fauna of the mountain, Dr. Tate found time also to make 515 collections of plants. These were deposited at the New York Botanical Garden, where they were studied and reported on by H. A. Gleason.” Albert S. Pinkus and P. S. Perberdy in 1938-39 visited Roraima by way of the Mazaruni River, British Guiana. Their collections of plants, representing some 292 numbers (collected and prepared almost entirely by their Arawak Indian assistant, Rufus Boyan), were studied and distributed by the New York Botanical Garden, where the first set is deposited. Sixteen new species were described from the Pinkus materials. No further plant collections of significance were made on Roraima until 1945 when Julian A. Steyermark visited the mountain. Dr. Steyermark’s first set is at the Chicago Museum of Natural History; duplicate sets are at the New York Bo- tanical Garden, the United States National Herbarium, and elsewhere. In earlier reports, precise geographical designation may not have been consid- ered significant in comparison to the overriding importance of initial exploration and investigation, since in the literature most of the collections of this region have been loosely assigned to British Guiana, whereas in point of fact most of these early collections are from Venezuela and Brazil. While many of the species involved may and probably actually do occur within the political limits of British Guiana, in great part there exists no actual record of them for that British colony. | Even as late as 1929, Gleason (p. 393) in his report on the Tate collection, wrote: **The small area on the summit of Mount Roraima is shared by Brazil, Venezuela, and British Guiana, and the dividing line between them has not been accurately located. All specimens in this article from the summit and slopes of the mountain have been credited to British Guiana as a matter of convenience.”’ Phelps® has discussed a similar confusion of geography as related to birds collected on various expeditions to Roraima, on which the birds were taken wholly “Bull. Torrey Club 56: 391. 1929. *Bull. Torrey Club 56: 339-390. 1929. *Phelps, William H. The geographical status of the birds éolledeesd at Mount Roraima, Bol. Soc. Venez. Ci. Nat. 83-95. 1939. reprint, Caracas, 1939. 90 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 within Venezuela but for the most part recorded as from British Guiana. He has pointed out that as delimited by natural watershed and boundary commission find- ings, more than five-eighths, the western and southern portion of the summit (and that part traversed by various expeditions) is Venezuelan, less than one-fourth, the northeastern part (which is designated largely ‘‘inaccessible’’ on the Commis- sion map), is British Guianan, and a very small portion, about one-eighth, on the southeastern portion Brazilian. Further botanical exploration of the Guayana Highland has been made without confusion of geography. For more than the past two decades the colleagues of Professor Henri Pittier, namely Captain Felix Cardona, Doctors Tobias Lasser, Francisco Tamayo, and others, have visited the Gran Sabana. The major collections of all of them are in the Venezuelan National Herbarium. Over the period October 1, 1928 to March 18, 1929, the Sidney F. Tyler—Ameri- can Museum of Natural History Expedition was led by Dr. Tate’ in the first ascent of Cerro Duida, on the western margin of the Guayana Highland. Tate’s important collections were deposited at the New York Botanical Garden, and became the basis of the voluminous and important ‘‘Botanical results of the Tyler-Duida Ex- pedition’’ by Gleason.°* Stimulated by the explorations of Roraima and Duida, The American Museum of Natural History, the American Geographical Society, and the New York Botan- ical Garden completed preliminary organization of an ambitious plan for the further over-all biological, geological, and geographical survey of the Guayana Highland. An outline entitled ‘Prospectus of the Pacaraima-Venezuela Expedition’’ was issued, detailing proposed objectives for the cooperative undertaking. Unfortu- nately, these well-balanced plans did not reach fruition. The American Museum was, however, able independently to carry out a comparable exploration under the sponsorship of Dr. William H. Phelps of Caracas. Tate’ led the important William H. Phelps-American Museum of Natural History Expedition of Auyan-tepui, near the Rio Caroni in the northeastern part of the Gran Sabana. Tate’s invaluable plant collections from the region for the third time came to the New York Botanical Garden. Report on them was made by Gleason and Killip.*° In the meanwhile, George S. Jenman, from 1903 to 1929 Government Botanist and Superintendent of the Botanic Garden, Georgetown, conducted extensive ex- ploration in British Guiana and collected on the Kaieteur Escarpment about Kaieteur Falls, on the Potaro River, which Im Thurn had visited in 1884. R. A. Alston, who was Assistant Government Botanist and Mycologist in British Guiana during 1923-1927, made two excursions into eastern Guayana. During August and September of 1925, he collected at Macreba Falls on the Kurupung River, where it drops from the sandstone plateau, and on the plateau itself along the Membaru trail. Collections made here represent some 130 numbers. In March, April, and May of 1926 Alston made an expedition up the Potaro River across the Kaieteur Pla- teau to the Ireng River. This trip netted about 98 collections. Both are chiefly deposited in the Jenman Herbarium, Georgetown, and at Kew. H. A. Gleason made Numerous important collections in the Potaro River Gorge in 1924. (Maguire vis- ited the magnificent Kaieteur Falls and exciting Kaieteur Escarpment as a stu- dent of zoology in 1925.) "Tate, G. H. H. & Hitchcock, C. B. Geogr. Rev. 20: 31-52. 1930. *Bull. Torrey Club 58: 277-506. 1931. °Geogr. Rev. 28: 452-473. 1938. °Brittonia 3: 141-204. 1939. 1953] BOTANY OF THE GUAYANA HIGHLAND 91 Cambridge University sent an expedition to the Kaieteur Savanna in 1933. T. G. Tutin has reported** on the botanical results in which sixteen new species were proposed. N. Y. Sandwith, botanist on the Oxford University Club Expedition to British Guiana in 1929, later visited the Kaieteur Savanna in August and Sep- tember of 1937. Many new species and records have come from his collections. In continuation of its program of research on the botany of the Guayana High- land, the New York Botanical Garden organized an expedition to the Kaieteur Es- carpment in 1944, Some twenty days were spent in this famous collecting area by Bassett Maguire in the company and under the guidance of D. B. Fanshawe, British Forest Officer. Later that same year, Maguire’? proceeded with the exploration of Tafelberg in central Surinam, the easternmost sandstone table mountain outlier of the Guayana Highland. The botanical results of the British Guiana field work (April 13=-May 23, 1944) and the Tafelberg Expedition (June 12—October 13, 1944) in Surinam have been published in a series of papers by him’* in conjunction with numerous collaborators. During 1944 and 1945 Julian A. Steyermark collected extensively in Ecuador and Venezuela. In the Guayana Highland he visited Duida and Roraima, and during the same period made the first botanical collections from Ptari-tepui in the State of Bolivar. Steyermark’s numerous collections form the basis of his important con- tributions’ to the flora of Venezuela, of which part 1 is at this writing already published. After the major expedition to Auyan-tepui, William H. Phelps and William H. Phelps, Jr. continued their thorough program in the study and collection of the avifauna- of Venezuela. Recent active field work in Guayana has largely been carried on by Mr. and Mrs. William H. Phelps, Jr. They have conducted a series of explorations of the sandstone tepuis of the Gran Sabana and Territorio Ama- zonas, beginning. with a visit to Mount Roraima, continuing successively with ex- peditions to Uaipan-tepui, Chimanta-tepui, Guaiquinima, Sipapo (P araque), Yavi,"* Part, and most recently with the ascent of Guanay and Camani in 1951. Mrs. Phelps (Kathleen D. Phelps) has been the botanist of these expeditions, and, often with the assistance of Charles B. Hitchcock, has made invaluable con- tributions to the botany of the Highland. The botanical materials of these expedi- tions are at the New York Botanical Garden, where collaborative studies and pub- lication on them are under way.’° Mrs. Phelps joined the staff of the New York Botanical Garden in 1950 as Collaborator in Venezuelan Botany. As a result, the first of the joint studies, an initial report on the piants of Uaipan-tepui, is at this writing in the hands of the editors.*’ In January and February of this year (1951) Charles B. Hitchcock, Gerald Budowski and Bassett Maguire had the pleasure of accompanying Mr. and Mrs. Phelps on their exploration of Cerros Guanay and Camani in the headwaters of the - Jour. Bot. 72: 306-314. f. 1-5. N 1934; 333-341. f. 6-11. D 1934, Geogr. Rev. 35: 563-579. 1945. *3Maguire, Bassett and Collaborators. Plant explorations in Guiana in 1944, chiefly to the Tafelberg and the Kaieteur Plateau. Bull. Torrey Club 75: 56-115, 182—230, 286-323, 374~438, 523~580, 633-671. 1948. *4Steyermark, Julian A. and Collaborators. Contributions to the flora of Venezuela. Fieldiana Bot. 28: 1-242. 1951. *SFor a narrative of the Cerro Yavi expedition, see: Charles B. Hitchcock. The Orinoco- Ventuari Region, Venezuela. Geogr. Rev. 37: 525~566. 1947. *SLasser, T. & Bassett Maguire. A Report on the Plants of the Phelps’ Cerro Yavi Expedition of 1947. Brittonia 7: 75-90. 1950. *7Maguire, Bassett & Phelps, Kathleen D. Botany of the Phelps’ ‘Venezuelan Guayana Se eedition—I. Uaipan-tepul, Estado Bolivar. Bol. Soc. Venez. Ci. Nat. lin press. 92 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 Rios Manapiare-Guaviarito, Territorio Amazonas. The first part of our report on the plants of these two sandstone mountains likewise is in the hands of the editors.” Upon the completion of the Kaieteur and Tafelberg report, the New York Bo- tanical Garden resumed its program of field excursions into the Guayana High- land. The first H. R. Kunhardt-New York Botanical Garden Expedition was con- ducted by Bassett Maguire, Louis Politi, and Bassett Maguire, Jr. to Cerro Sipapo (Paraque) in the Rio Cuao, tributary of the Upper Orinoco in Amazonas. Nearly three months were spent on its summit from October 1948 to February 1949. Immediately succeeding the first, the second Kunhardt-New York Botanical Garden Expedition was sent specifically in search of Arundinaria schomburgkii, which had not been recollected since its original discovery in 1839 by Robert Schomburgk on the southeastern tip of Cerro Marahuaca. Maguire and Maguire approached Marahuaca by way of the Cunucunuma, headwater stream of the Orinoco. On the way, three days were spent on the north summit of Cerro Duida. On May 9, 1949 a large community of Arundinaria, the famous curata from which the Maqui- ritare Indians make their blow-guns, was found along the base of Marahuaca’s enormous escarpment, high at 5500 feet altitude in the cloud forest of the south- east talus slope. In 1950 the New York Botanical Garden Expedition returned to Amazonas. Maguire, with Richard S. Cowan and John J. Wurdack, again spent eight days on the north summit of Duida, from November 18 to November 25. Tate and Hitch- cock had, from the South Escarpment, penetrated the mountain northward into the large central valley drained by Cano Negro. The 1950 expedition worked southward from its camp near the North Escarpment to the north ridges of Cano Negro. To the north of Duida, fifteen miles across the forested valley of the Rio Cunu- cunuma, is the completely escarpment-ringed Cerro Huachamacari, the ‘‘House of the Maquiritare Gods,’’ rising some 6500 feet above the valley floor. A tangential ledge (somewhat similar to that on the face of Roraima’s cliffs) traversing the 1500 foot South Escarpment, culminating in a 150 foot vertical erosion chimney, provided a means of ascent. Thirteen days, from Lecember 6 to December 18, were spent on the summit. Closely successive explorations were then made to the summits of Cerro Yapacana on the Orinoco (December 31 to January 6) and Cerro Moriche on the Ventuarjy (January 12 to January 21). Cowan and Wurdack extended the expedition of the New York Botanical Garden to the further exploration of the enormous Serrania Part from January 31 to Febru- ary 17, 1951. The party succeeded in reaching the summit repeatedly, the northern segment, Cerro Para, which is separated from the greater portion of the plateau- mountain by a deep canyon. From September 1947 through July 1948 Richard E.Schultes carried on explora- tion in the Upper Rio Negro drainage of Brazil and Colombia. On further extended exploration, he is in the same general region at the present writing. Schultes in large measure has retraced the routes of Richard Spruce, recollecting much of the Spruce materials and adding much new material of his own. In the course of his explorations, Schultes had the opportunity to visit the Cerros Campana and Chiri- biquete, neighboring sandstone mountains, both under 3000 feet in elevation, on the upper Apaporis in Vaupés, Colombia. The botanical results are being pub- *\facuire, Bassett & Phelps, Kathleen D. Botany of the Phelps’ Venezuelan Guayana Expedition—I. Territorio Amazonas. Bol. Soc. Venez. Ci. Nat. [in press. 1953] BOTANY OF THE GUAYANA HIGHLAND 93 lished by the explorer in a series of reports under the title ‘‘Plantae austro- americana.’’*” Along the eastern foothills of the Cordillera Oriental in Colombia is the ex- tensive Serrania Macarena, an enormous sandstone massif, possibly to be associ- ated geologically with the sandstone mountains of Guayana. Only recently it has been the object of a British Museum=Colombian Expedition, the results of which are as yet unpublished. W. R. Philipson, of the British Museum, was leader of botanical exploration. During the fall and winter of 1951=1952,’° the New York Botanical Garden will conduct expeditions to Cerro Guaiquinima in the State of Bolivar, in Vene- zuela, and with the British Forest Cepartment to the northeastern portion of the Pacaraima Mountains in British Guiana. During the 1952-1953 season, the New York Botanical Garden plans to complete this twenty-five year program in the exploration of Cerro Chimanta-tepui and outliers in the Gran Sabana. The further sustained field operation of the Servicio Botanico and the continuing exploration by Mr. and Mrs. Phelps may be expected to augment the collected materials and general knowledge of the region for a long time to come. It is hoped that the material and data so gained will have become sufficient for a preliminary floristic treatment of the Guayana Highland. Localities, collectors, dates, and exsiccatae numbers of the 1948-51 Venezuelan expeditions States of Anzoategui and Monagas. Bassett Maguire, H. R. Kunhardt, Louis Politi, October 21—November 4, 1948: 27207-27304. Cerro Sipapo (Paraque), Territorio Amazonas. Bassett Maguire, Louis Politi, Bassett Maguire, Jr. (after January 1), Novem- ber 15, 1948=March 15, 1949: 27318-29039. Cerro Duida, Cerro Marahuaca, Territorio Amazonas. Bassett Maguire and Bassett Maguire, Jr., April 18, 1949=May 21, 1949: 29040 —29221. Rio Atabapo, Territorio Amazonas. Bassett Maguire, October 17-October 26, 1950: 29222-29345. Cerros Duida, Huachamacari, Yapacana, and Moriche, Territorio Amazonas. Bassett Maguire, Richard S. Cowan and John J. Wurdack, November 2, 1950= January 21, 1951: 29346=31057. Cerro Part, Territorio Amazonas. Richard S. Cowan and John J. Wurdack, January 31, 1951=February 21, 1951: 31058~31599. Acknowledgements. Acknowledgement of assistance and services is always a. pleasant duty, but is regrettably one that never succeeds fully in meeting its genuine obligation to the many who contribute so much to successful exploration. Bot. Mus. Leafl. 15: 29—78. 1951. et praec. 2°Since this paper has been written, the 1951~1952 expeditions have returned to New York after having completed the explorations noted above, viz. the visits to Guaiquinima, and to the British Guiana—Pacaraima region. In addition Mr. and Mrs. William H. Phelps, Jr. have sponsored an excursion, conducted by Bassett Maguire, to Ilu-tepui, the north- emmost of the Mt. Roraima chain along the Venezuelan-British Guiana boundary. The botanical results of these latest expeditions will be reported in the course of the publica- tions of this series of papers. 94 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 To all, those directly mentioned and to many more unspecified contributors, we of- fer sincere and grateful acknowledgment. The prosecution of these explorations would not have been possible but for the facility lent by the many officials of the Government of Venezuela and the United States Department of State. Particular appreciation should be expressed to their Excellencies, the Ministers of Agriculture and Forestry, and of Information, to their Excellencies, Coroneles Paoli and Calcano, Governors of the Territorio Amazonas, to Dr. Tobias Lasser, Chief of the Botanical Service and Director of the Venezuelan National Herbarium, to Dr. A. Davila-Delgado, formerly Vene- zuelan Consul-General in New York, now Minister to Belgium, and to Dr. James H. Kempton, Agricultural Attaché to the American Embassy in Caracas. Mr. H. R. Kunhardt, chairman of the New York Botanical Garden Exploration Committee, retired president of the Venezuelan Petroleum Company, organized and sponsored the Kunhardt Expeditions to Cerros Sipapo, Duida, and Marahuaca. A major part of the following report derives from the collections of the Kunhardt Expeditions. Dr. William H. Phelps, who sponsored the Phelps -Auyan-tepui Expedition and Mr. William H. Phelps, Jr., both eminent ornithologists of Venezuela, continue to assist in the further program of exploration. Mrs. Phelps, Collaborator in Vene- zuelan Botany at the New York Botanical Garden, together with Mr. Charles B. Hitchcock, Secretary of the American Geographical Society, contributed in numer- ous ways to the botanical survey of Guayana. All have shared their rich experi- ences in exploration technique, supplied trained assistants, conducted inde- pendent and important plant field work, and have been ever generous hosts in Caracas and on their own effectively operated expeditions. In Ciudad Bolivar, Puerto Ayacucho, and San Fernando, numerous officials and friends have eased and simplified final organization at these interior river staging ports. Particularly in San Fernando Mr. William Northrup and Mr. Robert Shaylor have assisted in generous hospitality and use of boats and outboard motors. I shall ever be grateful to Mrs. Northrup who expertly nursed my son back to health after a severe attack of malaria and dysentery. Through officials of the Sinclair Oil and Refining Company, chiefly Mr. H. R. Kunhardt, Mr. E. T. Lincoln, Miss Julia Baumann, Mr. L. J. Maurovich, Mr. Adolph Michaelson, and Mr. George McKnight, ship passage to Venezuela has been pro- FIG. 1. Map showing the approximate location and distribution of sandstone mountains or plateaus that have been the subject of past exploration, or are the object of future ex- plorations anticipated by The New York Botanical Garden. Surinam *Ilu-tepui tCerro Jaua 1. Tafelberg 4. Auyan-tepui Venezuela, Territorio British Guiana oF Uaipan-tepui Amazonas 2. Kaieteur Plateau 6. Ptari-tepui 9. Cerro Yavi Venezuela, Estado 7. Chimanta-tepui (Acopan- 10. Cerro Yutajé Bolivar tepul ) 11. Cerro Guanay 3. Mt. Roraima 8. Cerro Guaigquinima 12. Cerro Camani 13. Cerro Sipapo (Paraque) 19. Cerro Marahuaca 14, Cerro Moriche ; Colombia, Comisaria Vaupés 15. Cerro Paru 20. Cerro Campana 16. Cerro Yapacana 21. Cerro Chiribiquete 17. Cerro Huachamacari Colombia, Intendencia Meta 18. Cerro Duida 22. Serrania de la Macarena *Ilu-tepui. Not designated on map, but lying 40 kilometers northwest of Mt. Roraima. Visited by the Phelps-New York Botanical Garden Expedition of 1952. tCerro Jaua. Not designated on map, but lying approximately 150 kilometers southwest of Cerro Guaiquinima. . 95 S VN. Ow 25 wi ay ae YY» a is eat eS a be Ps ie 9 pV ALAS Gos gos eR is: GW, On. Oss BOTANY OF THE GUAYANA HIGHLAND 1953] ‘ : 96 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 vided, and in Puerto la Cruz the facilities of the company’s staff house, commis- sary, and warehouse have been placed at our disposal. Through the generosity of Mr. Mack Lake, President of the Orinoco Mining Company, Mr. G. G. Lancaster has in a similar manner placed the Company’s facilities in Ciudad Bolivar at our disposal for the duration of the program of exploration. ) Special funds, supplies and equipment have been generously donated by many friends of exploration and the New York Botanical Garden. Chief among these should be acknowledged the financial assistance of the Garden Club of America; the contributions of trade goods, miscellaneous supplies, hunting and fishing equipment by Sears Roebuck and Company, and by the Johnson Motor Com- pany a 1951 Johnson ‘‘Seahorse’’ twenty-five horse-power outboard motor, so es- sential to travel on the ‘‘highways’”’ of the interior, Venezuela’s great rivers. New drugs and medicines developed within recent years have made the work of tropi- cal exploration comparatively safe; Parke, Davis and Company supplied us with chloromycetin and camoquin, Charles Pfizer and Company provided terramyacin and penicillin, and Squibb and Company donated penicillin and sulfadiazine. The forerunners in this program, Dr. G. G. H. Tate, Dr. H. A. Gleason, Dr. Tobias Lasser, and the many collaborators who have contributed their talent and knowledge to the earlier reports on the explorations of Roraima, Duida, Auyan- tepui, Kaieteur, Tafelberg, and Yavi; Dr. Henri Pittier and his associates, Dr. Tobias Lasser, Dr. Francisco Tomayo, Dr. L. Schnee, Captain Felix Cardona; and Dr. Julian A. Steyermark, who has conducted independent exploration, have provided a background of immeasurable consequence. And, finally, in the present assessment of collections, the experienced judgment of colleagues is being sought to assure the most competent understanding and interpretation of the flora of Guayana that is possible at this time. The report of each contributor will ap- pear over his own name. : THE NEW YORK BOTANICAL GARDEN NEW YORK 1953] BOTANY OF THE GUAYANA HIGHLAND 97 ERIOCAUL ACEAE*! Eriocaulon atabapense Moldenke. In moist sand among rocks 15 km. above San eye ees Rio Atabapo, Rio Ori- noco, Maguire 29256. The species is known only from this and three other collec- tions: Williams 13858, 14084 from the Rio Atabapo, and a specimen without data in the Herbario Nacional de Venezuela which is probably a duplicate of one of the Williams collections. A. C. Smith 2280, cited as this species in Phytologia 3: 182 (1949), proves to be E. tenutfolium. Eriocaulon humboldtii Kunth. Occasional herb in wet border of Savanna III, 125 m. alt., Cerro Yapacana, Rio Orinoco, Maguire, Cowan, & Wurdack 30464. The species ranges from Colombia (Méta) and Venezuela (Amazonas, Anzoategui, Bolivar, and Nelta Amacuro) to British Guiana and Brazil (Amazonas and Mattogrosso). Eriocaulon tenifolium Klotzsch. With chalky-white heads, frequent in wet savanna northwest of the base of Cerro Moriche, 150 m. alt., Rio Ventuari, Maguire, Cowan, & Wurdack 30984. The species is represented also by Schomburgk 285 and 448 and A. C. Smith 2280 from British Guiana. It has been reported from Bolivar, Venezuela, and from Amazonas and Bahia, 3razil, but perhaps erroneously. Paepalanthus capillaceus var. proliferus Gleason. Submerged aquatic perennial herb, with heads emergent white, becoming brown, abundant in the bed of Culebra Creek, 1200 m. alt., Cerro Duida, Rio Cunucunuma, Maguire, Cowan, & Wurdack 29611; locally common in bed of Cafho Culebra, 1300 m. alt., Cerro Duida, Maguire & Maguire 29153. Only some of the heads on these specimens have the proliferations described by Gleason; however all the heads show the characteristic pilosity not seen on any of the specimens of the typical form of the species thus far examined by me. The variety is known only from Cerro Duida (Steyermark 58138) and, in British Guiana, from Mount Roraima (Tate 263, 552) and the Kaieteur Plateau (Maguire & Fanshawe 23243). Paepalanthus fasciculatus (Rottb.) Korn. Frequent on white sand, savanna in the vicinity of Base Camp, 150 m. alt., Cerro Sipapo (Paraque), Maguire & Politi 28309. A widely distributed and very variable species known from Colombia (Cundinamarca, Meta, Santader del Norte, and Vaupeés), Venezuela (Amazonas and Bolivar), British Guiana, Surinam, French Guiana, and Brazil (Amazonas and Para). Paepalanthus kunhardtii Moldenke, sp. nov. Herba parva caulescens; caulibus brevibus brachiatis; foliis rosulatis oblongo- Ovatis erectis utringue lucidis 2.5-4 cm. longis, 3-5 mm. latis, ad basim longe villosis subcoriaceis; vaginis laxiusculis 4-5 cm. longis multistriatis glabris, lamina 5-G mm. longa erecta longe ciliata; pedunculis graciliusculis brunneis © 19-23 cm. longis tricostatis dense laxeque pilosis; capitulis hemisphericis atro- gtiseis 5-8 mm. diametro. Small caulescent herb; stems 1.5-3 cm. long, often short-branched, densely long-villous; leaves rosulate, those at the tip of the main stem and of the branches forming what appears to be one many-leaved rosette, oblong-ovate, bright green and shiny on both surfaces, erect, 2.5-4 cm. long, 3-5 mm. wide at the mid-point, ampliate-clasping at the base, glabrous on both surfaces except for the long- villous base, acute or subacute at the apex, thick-textured or subcoriaceous; 71By H. N. Moldenke. ‘ - 98 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 sheaths rather loose, 4-5 cm. long, many-striate, slightly twisted, glabrous except for the mouth, obliquely split at the apex, the blade firm, erect, 5-6 mm. long, long-ciliate on the lower half of the margins; peduncles rather slender, 1 or 2 per stem and branch, 2-7 per plant, brownish, 19-23 cm. long, 3-costate, slightly twisted, densely loose-pilose with whitish ascending hairs; heads hemispheric, dark-cray, 5-& mm. in diameter; involucral bractlets elliptic-ovate, dark brown or blackish, about 2 mm. long and 1 mm. wide, obtuse or subacute at the apex, glabrous or minutely ciliolate at the apex; receptacle long-pilose; receptacular bractlets obovate-spatulate, dark-brown, about 2.5 mm. long and 0.8 mm. wide, obtuse and densely white-barbellate at the apex; staminate florets: sepals 3, sep- arate, oblanceolate, brownish, about 2 mm. long and 0.5 mm. wide, obtuse and densely white-barbellate at the apex, otherwise glabrous; petals 3, connate into a hyaline infundibular tube about 1.5 mm. long, glabrous; stamens 3, exserted, white; anthers white; rudimentary pistil brown, about 0.7 mm. long, 3-parted, in- cluded; pistillate florets: sepals 3, elliptic, about 2.5 mm. long and 0.5 mm. wide, brownish, obtuse, densely white-barbellate at the apex, lightly pilosulous on the back; petals 3, hyaline, linear, about 2.5 mm. long and 0.2 mm. wide, subacute, densely long-villous, the basal white hairs longer than the petal, the upper ones gradually shorter so that all reach about the same level; pistil brown, compara- tively stout, about 2.7 mm. long, glabrous; style about 0.5 mm. long; stigmas 3, about 1 mm. long; style-appendages 3, about 1.5 mm. long; ovary small, about 0.5 mm. long and wide. TYPE: in wet sphagnum hummocks, Camp Savanna, 4500 feet alt., Cerro Si- papo (Paraque), Amazonas, Venezuela, December 11, 1948, B. Maguire & L. Politi 27588; New York Botanical Garden. Paepalanthus lamarckii Kunth. Heads grayish-white, occasional on moss-covered boulders,rapids above Playa Alta, Rio Cunucunuma, Rio Orinoco, Maguire, Cowan, & Wurdack 29499. This species has the widest and most amazing distribution of any member of the fame ily, having been collected in British Honduras, Panama (Coclé), Cuba (Pinar del Rio), Hispaniola, Isla de Pinos, Trinidad, Venezuela (Amazonas, Bolivar, Falcon, and Monagas), British Guiana, Surinam, French Guiana, Brazil (Amazonas, Ceara, Goyaz, Maranhao, Para, Pernambuco, and Piauhy), and Marajo Island; also in French Guinea, Sierra Leone, French Equatorial Africa (Gabun), Tanganyika Ter- ritory, and Madagascar! Paepalanthus maguirei Moldenke. With white flowers, common among rocks, Culebra Rapids I, Rio Cunucunuma, Rio Orinoco, Maguire, Cowan, & Wurdack 30364. The species is also known from Tafelberg, Surinam (Maguire 24241, 24832). Paepalanthus perplexans var. wurdacki Moldenke, var. nov. Haec varietas a forma typica speciei recedit foliis 1.5-2.5 cm. longis, 2-2. a mm. latis, capitulis subglobosis 4-5 mm. diametro. TYPE: on cumbre, frequent, 2000 m. alt., Serrania Part, Rio Part, Cafio Asisa Rio Ventuari, Amazonas, Venezuela, February 2, 1951, R. S. Cowan & J. J. Wur- dack 31141; New York Botanical Garden. The typical form of the species is known only from Ptari-tepui, Bolivar. Paepalanthus tatei Moldenke. Frequent in bogs about pool, Camp Savanna, 4500 feet alt., Cerro Sipapo (Paraque), Maguire & Politi 27702; frequent in moist sandy banks among rocks, Danta Falls, Rio Cuao, Rio Orinoco, Maguire & Politi 27343; among rocks in open © 1953] BOTANY OF THE GUAYANA HIGHLAND 99 stream bed, Culebra Creek, 1300 m. alt., Cerro Duida, Rio Cunucunuma, Maguire, Cowan & Wurdack 29631. The species is also known from Cerro Guanay, Cerro Sipapo, and Cerro Yavi, as well as from Auyan-tepui in Bolivar, from the state of Lara, and from the Sierra de la Macarena in Colombia. Paepalanthus williamsii Moldenke. Occasional herb in moist places around border of Savanna No. III, 125 m. alt., Cerro Yapacana, Rio Orinoco, Maguire, Cowan & Wurdack 30806; infrequent in wet places, border of Savanna No. III, 125 m. alt., Maguire, Cowan & Wurdack 30463. The species was hitherto known only from the Savanna de San Antonio, also in Amazonas (Williams 15051) and from a recent collection in the state of Amazonas, Brazil. Syngonanthus alleni var. parvus Moldenke, var. nov. Haec varietas a forma typica speciei recedit caulibus parvioribus, foliis gla- bris attenuatis, bracteis paucis brevioribus, pedunculis paucioribus, et capitulis florisque minoribus. TYPE: on shallow wet sand on rock outcrop behind (east of) Hotel Amazonas, Puerto Ayacucho, Amazonas, Venezuela, October 24, 1950, B. Maguire, R. S. Cowan & J. J. Wurdack 29238; New York Botanical Garden. The typical form of the species is known only from Vaupes, Colombia. Syngonanthus anomalus (K6rn.) Ruhl. Stamens white, frequent on marshy bank along river, 1 km. above Culebra Rap- ids, Rio Cunucunuma, Rio Orinoco, Maguire, Cowan & Wurdack 30409. This spe- cies is known from Amazonas and Bolivar, Venezuela, as well as from British Guiana and from Amazonas, Brazil. Several rather poorly defined varieties and forms have been described. Syngonanthus biformis (N. F. Br.) Gleason. Common on mossy moist sandy banks along rocks, Danta Falls, 150 m. alt., Rio Cuao, Rio Orinoco, Maguire & Politi 27342; with white flowers, along stream in rain forest, Cafio Culebra, 1000-1100 m. alt., Cerro Duida, Rio Cunucunuma, Maguire, Cowan & Wurdack 29518. The species is known from Colombia (Vaupés) and Venezuela (Amazonas, Bolivar, and Sucre) to British Guiana and Surinam. Syngonanthus caulescens (Poir.) Ruhl. Moist sandy banks over rocks at Danta Falls, 460 feet alt., Cerro Sipapo (Paraque), Maguire & Politi 27342a; frequent in moist sandy banks among rocks, Danta Falls, Rio Cuao, Rio Orinoco, Maguire & Politi 27343a; infrequent along sandy stream banks under Mauritia, 18 km. south of Santa Barbara, Monagas, Maguire, Kunhardt & Politi 27299. The species is widely distributed in the New World, from Mexico (Veracruz) through Costa Rica, Colombia (Cundinamarca, Meta, Santander del Norte, and Vaupés), Venezuela (Amazonas, Anzoategui, Bolivar, Carabobo, and Monagas), British Guiana, Surinam, and French Guiana to Peru (San Martin), Brazil (Amapa, Amazonas, Bahia, Goyaz, Mattogrosso, Minas Geraes, Parana, Pernambuco, Piauhy, Rio de Janeiro, Rio Grande do Sul, and Sao Paulo), Bolivia (Santa Cruz), Paraguay, and Argentina (Corrientes and Misiones). Syngonanthus cowani Moldenke, sp. nov. Herba parva caulescens 2-8 cm. alta; caulibus simplicibus vel usque ad 3- aggregatis gracilibus rigidis erectis atrobrunneis vel nigrescentibus marginatis densiuscule laxeque pilosis; foliis caulinis verticillatis adscendente-erectis; foliis basalibus numerosissimis dense rosulatis recurvo-reflexis anguste lineari- bus rigidiusculis 8-12 mm. longis subulatis glabratis l-nervatis; inflorescentiis ‘ 100 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 umbellatis densissimis terminalibus hemisphaericis vel subglobosis 1-2 cm. dia- metro; pedunculis 20-50 filiformibus 2-8 mm. longis parce pilosulis vel glabris. Small caulescent herb, 2-8 cm. tall; stems solitary or sometimes with 1 or 2 smaller ones in addition, the main ones slender, rigid, erect, dark brown or black- ish, more or less flattened and angle-margined, rather densely and loosely pilose with gray, irregular, mostly more or less appressed hairs, bearing a whorl of many cauline leaves 0.8-3.5 cm. from the base, the side stems without the cauline whorl of leaves; basal leaves very numerous, densely rosulate, recurved-reflexed to the ground, narrowly linear, rather rigid, 8-12 mm. long, less than 0.5 mm. wide, subulate-tipped, glabrate, l-nerved, not fenestrate; cauline leaves similar but ascending-erect; inflorescence a very dense terminal umbel which is hem- ispheric or almost subglobose, 1-2 cm. in diameter on the main stems, 6-8 mm. in diameter on the secondary stems, composed of 20=50 or more short pedunculate heads, each peduncle subtended by a leaf-like bract; peduncles filiform, 2-8 mm. long, sparsely pilosulous or glabrous; sheaths absent, represented by an equitant- navicular bract to about 7.5 mm. long, minutely pilosulous, acute at the apex; heads obconic or subhemispheric, 3-4 mm. in diameter; receptacle short-pilose; involucral and receptacular bractlets hyaline or whitish, elliptic, about 2 mm. long and 0.5 mm. wide, acute at the apex, sparsely pilosulous on the back; stamin- ate florets: sepals 3, whitish-subhyaline, connate at the base, the free portions elliptic, 1-1.5 mm. long, subacute or rounded at the apex, often with 1 or a few unicellular hairs at the tip; petals 3, white, connate into a slender infundibular tube for 1.5-2 mm., glabrous, the terminal free portion elliptic-spatulate, about 1 mm. long, rounded at the apex, glabrous; stamens 3, borne on the petals at the apex of the tube, included; filaments filiform, 0.4-0.5 mm. long, glabrous, white; anthers white; rudimentary pistil reaching the base of the free filaments; stigmas 3, white-capitate; pistillate florets: sepals 3, separate, whitish, elliptic, about 1.5 mm. long and 0.5 mm. wide, acute at the apex, sparsely pilose on the back; petals 3, white, spatulate, connate at the middle, free at base and apex, about 2 mm. long, to 0.4 mm. wide, rounded at the apex, glabrous; style about 0.5 mm. long, glabrous; stigmas 3, about 0.5 mm. long, white-capitate; ovary narrow- elliptic, glabrous, 3-sulcate, 3-celled, 3-ovulate. TYPE: on sabanita 500 m. southeast of Savanna III, 125 m. alt., Cerro Yapa- canna, Rio Orinoco, Amazonas, Venezuela, December 31, 1950, B. Maguire, R. S. Cowan & J. J. Wurdack 30466; New York Botanical Garden. Another collection is Maguire, Cowan & Wurdack 30780, locally frequent on Savanna No. I, 125 m. alt. Syngonanthus flavipes Moldenke, sp. nov. Herba acaulescens; foliis caespitosis lineari-oblongis crassis 2.5~4.5 cm. longis patentibus vel circinato-incurvis obtusiusculis utringue minutissime pulverulento-puberulis vel glabris; vaginis tenuissimis arcte adpressis 4.5=7 cm. longis obscure multistriatis glabrescentibus, lamina lanceolata erecta arcte ad- pressa scariosa-marginata glabra vel ad apicem minute pilosula; pedunculis nu- merosissimis flavis 33-45 cm. longis 3-costatis glabris lucidis; capitulis hem- isphaericis albis 3-6 mm. diametro. > Acaulescent herb; leaves cespitose, linear-oblong, thick-textured, 2.5=4.5 cm. long, about 1.5 mm. wide, spreading or the inner ones often circinately in- curved, rather blunt at the apex, several-nerved beneath, not fenestrate, very minutely and obscurely pulverulent-puberulent on both surfaces or glabrate; sheaths very slender, closely appressed, far surpassing the leaves, 4.5-7 cm. long, obscurely .many-striate, somewhat twisted, glabrescent, obliquely split at the apex, the blade lanceolate, erect, closely appressed, scarious-margined, glabrate or very minutely pilosulous at the acute apex; peduncles very numerous, 1953] BOTANY OF THE GUAYANA HIGHLAND 101 yellow, about 30 per plant, 33-45 cm. long, 3-costate, glabrous, shiny; heads hemispheric, white, 3-G mm. in diameter; involucral bractlets very tough and firm, concave, closely appressed to the head, yellowish-stramineous, elliptic, about 2.5 mm. long and 1 mm. wide, rounded at the apex, glabrous, very shiny; re- ceptacle long-villous; receptacular bractlets whitish-subhyaline, spatulate- obovate, about 2.5 mm. long and 0.6-1 mm. wide, rounded at the apex, appressed- pilose on the entire back; staminate florets: sepals 3, separate almost to the base, spatulate, whitish-subhyaline, about 2 mm. long and 0.5 mm. wide, rounded at the apex, densely barbellate-pilose on the back above the middle; petals 3, connate into a subhyaline infundibular tube about 1.7 mm. long, glabrous; stamens 3; pis- tillate florets: sepals 3, hyaline, narrow-elliptic, about 2 mm. long and 0.5 mm. wide, very shortly barbellate on the back below the apex, otherwise glabrous; petals 3, narrowly oblong-spatulate, about 1.8 mm. long and 0.3 mm. wide, gla- brous; style about 1 mm. long, glabrous; stigmas 3; ovary subglobose, about 1 mm. long, 3-sulcate. TYPE: in wet places, Savanna No. III, 125 m. alt., Cerro Yapacana, Rio Ori- noco, Amazonas, Venezuela, December 31, 1950, B. Maguire, R. S. Cowan & J. J. Wurdack 30465; New York Botanical Garden. Syngonanthus gracilis (Kérn.) Ruhl. Locally frequent in moist sand among rocks, 15 km. above San Fernando, Rio Atabapo, Rio Orinoco, Maguire 29267. This is a widespread and extremely vari- able species. The typical form is known from Colombia (Meta and Vaupés), Vene- zuela (Amazonas, Bolivar, and Sucre), British Guiana, and Surinam to Brazil (Amazonas, Minas Geraes, Para, and Piauhy) and Uruguay. No less than 16 poorly defined and poorly understood varieties have been described, chiefly from Brazil. Syngonanthus gracilis var. glabriusculus Ruhl. With white flowers, in Vellozia association, cumbre at 1200 m. alt., Cerro Yapacana, Rio Orinoco, Maguire, Cowan & Wurdack 30714; frequent on open banks, lower Cafio Negro, 4400 feet alt., Rio Cuao, Rio Orinoco, Maguire & Politi 27917. This variety is also known from Bahia and Minas Geraes, Brazil. Syngonanthus humboldtii var. elongatus Moldenke, var. nov. Haec varietas a forma typica speciei recedit ramis 30-50 cm. longis, verticel- lis foliorum 14-21, et ramis saepe brachiatis. TYPE: on Savanna No. III, 125 m. alt., Cerro Yapacana, Rio Orinoco, Ama- zonas, Venezuela, January 1, 1951, B. Maguire, R. S. Cowan & J. J. Wurdack 30558. New York Botanical Garden. Syngonanthus humboldtii var. glandulousus Gleason. Dominant savanna herb, common in wet savanna northwest of base of Cerro Moriche, 150 m. alt., Rio Ventuari, Maguire, Cowan & Wurdack 30985. The variety is known also from the lowland savannas about Cerro Duida (Steyermark 57857, Tate 315), Auyan-tepui (Tate 1308), and San Fernando de Atabapo (Holt & Gebriger 234). Syngonanthus humboldtii yar. macrocephalus Moldenke, var. nov. Haec varietas a forma typica speciei recedit capitulis 7-9 mm. latis et pedun- culis densiuscule albido-pilosis, pilis irregulariter adpressis saepe contortis non glanduliferis. TYPE: in depressions in rocks on the southeast slopes of North Mountain, 5000 to 6000 feet alt., Cerro Sipapo (Paraque), Amazonas, Venezuela, December 12, 1948, B. Maguire & L. Politi 27649; New York Botanical Garden. A second collection is Maguire & Politi 27796 from bogs near the summit of West Peak, 5300 feet alt. ‘. 102 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 Syngonanthus humboldtii var. orinocensis Moldenke, var. nov. Haec varietas a forma typica speciei recedit caulibus crassioribus, foliis basalibus 3-8 cm. longis, foliis caulinis usque ad 2.3 cm. longis, pilis glanduli- feris. TYPE: under thickets on moist white sand about borders of small ‘‘laja,’’ Rio Temi, one hour below Yavita, Rio Atabapo, Rio Orinoco, October 20, 1950, B. Maguire 29340; New York Botanical Garden. Syngonanthus longipes Gleason. Annuals to 1 m. tall, occasional in little wet savanna, 150 m. alt., northwest of the base of Cerro Moriche, Rio Ventuari, Maguire, Cowan & Wurdack 30983. The species was hitherto known only from the Mount Roraima region of British Guiana (Appun 1199, ImThurn 33, Schomburgk 1060) and, in Bolivar, Venezuela, from 1100 m. on Mount AuyAn-tepui (Tate 1329) ‘and from 1220 m. between Santa Teresita de Kavanayén and the base of Ptari-tepui (Steyermark 60304). Syngonanthus oblongus (K6rn.) Ruhl. With pale green leaves and white heads, at base of waterfalls, Culebra Creek, 1300 m. alt., Cerro Duida, Rio Cunucunuma, Maguire, Cowan & Wurdack 29630; occasional in moist places among rocks, Cafio Culebra, 1300 m. alt., Cerro Duida, Maguire & Maguire 29158. Whether var. aequinoctialis Ruhl. is really distinct is questionable. The characters given by Ruhland do not seem to be constant, but the typical form of the species seems to have macroscopically plainly villous heads, while the variety has them subglabrous. On this basis the specimens cited above represent the typical species which was known hitherto only from Colombia (Vaupés) and Brazil (Amazonas, Bahia, Goyaz, Mattogrosso, and Piauhy). Syngonanthus phelpsae Moldenke. Infrequent along Cafio Grande, Campo Grande, 4500 feet alt., Maguire & Politi 27584; frequent in sphagnum hummocks, wet shrub savanna, Cafo Negro, 3500 feet alt., Maguire & Politi 27697. Syngonanthus phelpsae var. elongatus Moldenke, var. nov. Haec varietas a forma typica speciei recedit caulibus 5 cm. elongatis et foliis 2.5~4 cm. longis. TYPE: on cumbre at west rim, 2000 m. alt., Cerro Part, Amazonas, Venezuela, January 31, 1951, R. S. Cowan & J. J. Wurdack 31098; New York Botanical Gar- den. The collectors note that all the material of this number was taken from a single glump. Syngonanthus simplex (Miq.) Ruhl. Frequent on white sand, border of small savanna in the vicinity of Base Camp, 160 m. alt., Cerro Sipapo (Paraque), Maguire & Politi 28035. The species is known from Amazonas, Bolivar, and Falcon, in Venezuela, as well as from British Guiana, Surinam, and Brazil (Minas Geraes). Syngonanthus tenuis (H.B.K.) Ruhl. Infrequent in wet places, Savanna No. III, 125 m. alt., Cerro Yapacana, Rio Orinoco, Maguire, Cowan & Wurdack 30473. The species is known only from Venezuela (Amazonas and Bolivar) and Brazil (Amazonas and Para). Syngonanthus yapacanensis Moldenke, sp. nov. | Herba brevissime caulescens: foliis linearibus 1-2 cm. longis, ca. 0.5 mm. — latis, patentibus brunnescentibus obtusis glabris firmis; vaginis pertenuibus arcte adpressis 1-1.5 cm. longis glabris; pedunculis paucis gracillimis stramineis 3.5- 9.5 cm. longis tricostatis basim versus plusminusve albo-pilosis, apicem versus glabrescentibus; capitulis hemisphaericis albis 3-6 mm. latis. 1953] BOTANY OF THE GUAYANA HIGHLAND 103 Very shortly caulescent herb; stems 0.5~2 cm. long, or perhaps longer, rooting near the apex, often branched and leaf-bearing there, the leaf-bearing portion usually about 5 mm. long; leaves linear, 1~2 cm. long, about 0.5 mm. wide, spreading, brunnescent in age, blunt at the apex, glabrous, rather firm- textured, not plainly venose; sheaths very slender, closely appressed, 1-1.5 cm. long, not plainly striate nor twisted, glabrous, obliquely split at the apex, the blade oblong, about 3 mm. long, erect, blunt at apex; peduncles 1 or 2 per branch, very slender, stramineous, 3.5-9.5 cm. long, 3-costate, more or less white-pilose toward the base, glabrescent toward the apex; heads hemispheric, white, 3-6 mm. wide; involucral bractlets in several series, the outermost smallest, the innermost largest, stramineous, elliptic-obovate or the outermost ovate, 1.6-2.4 mm. long, 1-1.5 mm. wide, rounded and usually emarginate-split at the apex, the outermost often more or less laciniate, glabrous, shiny; receptacle densely long-pilose; receptacular bractlets oblong, whitish, about 1.5 mm. long and 0.5 mm. wide, ob- tuse at the apex, very densely long-barbellate with white hairs on the back at the apex, otherwise glabrous; staminate florets: sepals 3, separate, stramineous, spatulate, about 1.5 mm. long and 0.5 mm. wide, rounded and rather densely barbellate at the apex, otherwise glabrous; petals 3, connate into a stramineous infundibular tube about 1.5 mm. long, glabrous; stamens 3; pistillate florets: sepals 3, pale-stramineous, oblong, about 2.2 mm. long and 0.4 mm. wide, sparsely short-pilose at the apex on the back, otherwise glabrous; petals 3, subhyaline, lanceolate, about 2 mm. long, attenuate at the apex, sparsely short-pilose at the apex, otherwise glabrous; style about 1 mm. long, glabrous; stigmas 3, about 0.5 mim. long; ovary stramineous, subglobose, about 0.8 mm. long and wide, 3-sulcate, 3-celled, 3-ovulate. TYPE: locally frequent in wet savanna No. I, 125 m. alt., Cerro Yapacana, Rio Mrinoco, Amazonas, Venezuela, January 7, 1951, B. Maguire, R. S. Cowan & J. J. Wurdack 30782; New York Botanical Garden. LEGUMINOSAE-CAESALPINIOIDEAE In attempting to identify the collections of the genus Aldina it was at once apparent that one of the collections was most certainly a new species and further study revealed that the other three represented material of a second undescribed species. With the exception of two new species described by Ducke nothing has been added to our understanding of this interesting genus since the treatment in the Flora brasiliensis; here no key was presented and the descriptions were often incomplete and inadequate. Because of this situation, I have undertaken a prelim- inary review of the genus. The material on which this review is based is admittedly scanty and for this reason it is to be considered only as a preliminary study. It was a fortunate cir- cumstance to have in our herbarium the holotypes or isotypes of all but one of the described species and this one was kindly loaned by the U. S. National Her- barium; for this assistance I am indeed grateful. Key to the Species and Varieties of Aldina 1. Leaflets completely glabrous on lower surface as well as the upper. ........4.- 2 1. Leaflets pubescent on lower surface, sometimes obviously but often minutely so. . 3 2. Mature buds 7-8 mm. long, petals 7-9 mm. long; stipe of ovary as long as or shorter than ovary; leaflets broadly ovate to lanceolate; inflorescence much-branched. ame asta st aly of 6 es a five ole es es ben ete 6 Se - 10. A. heterophylla. 2. Mature buds 15 mm. long, petals 15-20 mm. long; stipe of ovary 2-3 times as long as ovary; leaflets oblong, oblong-oval, or oblong-ovate; inflorescence racemose or lax ESS CGC ol: Ah A 4a. A. latifolia var. latifolia. ‘ 7 104 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 2 3. All parts except upper leaf surface fuscous-velvety; calyx usually split into two parts. dy Tpiceien Arse inne “eying inne o whi Ma abel ace sl Ml teal ea oes 3. A, kunhardtiana. 3. Pubescence where present minute, densely and closely-appressed, golden or golden- ~ brown; calyx split into 5—5 irregular pafts. .%.'.-« so. « =.« «10/aeise)s ae ee 4 ‘Leaves “mostly 7=9=1 1-folvolaté. 00s \s 2 ss so seo a hee ee oe SP, goa eae ee ot as 4, Leaves usually not over 5-foliolate (to 7-foliolate in A. occidentalis), ........+. 6 5. Leaflets 3-4 cm. wide, lanceolate-oblong, apices long-acuminate; mature buds about 1.5 cm. long; petals oblong, 2 cm. long, 0.8 cm. wide; filaments about 1.5 cm. long; ovary 5 mm. long with a stipe up to 10 mm. long. ~....+...c..5-. 2. A. polyphylla. 5. Leaflets (5.5—)7.5-11 cm. wide, oblong or oblong-oval, apex abruptly short-acuminate or acute; mature buds 2-2.5 cm. long; petals obovate-orbicular with cuneate base, 3-4.5 cm. long, 2.5-3.5 cm. wide; filaments 2.5-3.5 cm. long; ovary 7-8 mm. long | (sometimes two ovaries produced per flower), stipe 1.5-1.8 cm. long........... 3 ae Tee ee re aE ee ee la, .A. insignis var. insignis. 6. Hairs on lower leaflet surface strictly appressed, directed toward the apex and margin | of leaflet, apex of leaflets not retuse or rotund. ...s.s2 cee eecebes ‘oe 6. Hairs on lower leaflet surface erect or suberect, collapsed and more or less decumbent on.drying, apex.of leaflets rotund aad retuse... cw, csi a mine ope ee eee I 7. Leaflets 7-8.5 cm. long, 3.5—4.5 cm. wide, lower leaflet surface obscurely pruinose- ( lepidote; mature buds about 1 cm. long, pecals 2 cm. or less in length, stipe of ovary . glabrous,’ about T mm: long, style :glabrous. si)'y .a°s s-c0 bs oe 9. A. occidentalis. 7. Leaflets 10-13 cm. long, 5.5-6.5 cm. wide, lower beetles. surface evenly glaucous; ma- ture buds about 2.5 cm. long, petals aie 4.5 cm. long, stipe densely pubescent, 6-9 mm. long, style sparsely gray-sericeous. ....2...-. .. lb, A. insignis var. retusa. 8. Ovary glabtous or subglabrous.’ 2... 200 bce w se 8 ws © 6 a ooo oe ee ee 9 9... Ovary dénsély pubescent. Ors os be ees ee Me ae ss Be oe eee : © ab ee ee ee 9. Blade of leaflets 21-22 cm. long, oblong, with 12-16 pairs of primary veins; inflores- cence a sparsely-branched panicle, 20-30 cm. long, petals oblanceolate, 20-25 mm. long,. stipe. of ovary nearly as-lome as OVARY sé a * My a Me ‘ . r 7 ny aa omic - LS I - - rR ee aes a= ay —~ ~-- PUBLICATIONS OF THE NEW YORK BOTANICAL GARDEN Flora of Bermuda, by Nathaniel Lord Britton. 585 pages, 519 figures. 1918. $3.50. The Bahama Flora, by Nathaniel Lord Britton and Charles Frederick Millspaugh. 695 pages. 1920. $6.25. Flora of the Prairies and Plains of Central North America, by P. A. Rydberg. 969 pages, 601 figures. 1932. $6.00. Plants of the Vicinity of New York, by Henry Allan Gleason. 284 pages, illustrated. 1935. $2.00. The New Britton and Brown Illustrated Flora of the Northeastern United States and Adjacent Canada, by Henry Allan Gleason. Including descriptions and illustrations of 4660 species. Three volumes, boxed. $30.00. North American Cariceae, by Kenneth Kent Mackenzie. 539 figures of Carex and re- lated species by Harry C. Creutzburg, with a description of each. Two volumes. 1940. $17.50. Keys to the North American Species of Carex, by Kenneth Kent Mackenzie. From North American Flora, Vol. 18. 1941. $1.25. Vegetables and Fruits for the Home Garden. Four authoritative articles reprinted from the Journal of The New York Botanical Garden, edited by Carol H. Woodward. 21 pages, illustrated. 1941. 15 cents. Food and Drug Plants of the North American Indian, by Marion A. and G. L. Wit- trock. Two articles, illustrated, from the Journal of The New York Botanical Garden. 1942. - 15 cents. Catalog of Hardy Trees and Shrubs. A list of the woody plants in cultivation outdoors at The New York Botanical Garden in 1942. 127 pages, map and 20 illustrations. 1943. 75 cents. Succulent Plants of New and Old World Deserts, by E. J. Alexander. 350 species, 100 illustrated. 64 pages. Third edition, 1950. Bound in paper, 50 cents. The Flora of the Unicorn Tapestries, by E. J. Alexander and Carol H. Woodward. 28 pages, illustrated. Third edition, 1950. Bound in paper, 25 cents. Periodicals Contributions from The New York Botanical Garden. Technical papers written by students and members of the staff and reprinted from various journals. Fourteen volumes. 1899-1913. 25 cents each, $5.00 a volume. Memoirs of The New York Botanical Garden. Issued irregularly. 1900- Vols. 1-6, to members of the Garden, $1.50 a volume; to others, $3.00. Vol. 7, to members, $2.50; to others, $5.00. Vol. 8, $10.00. Journal of The New York Botanical Garden. Notes, news, and non-technical articles on botany and horticulture, illustrated. (Replaced by The Garden Journal of The New York Botanical Garden.) 51 volumes. 1900-1950. North American Flora. Descriptions of the wild plants of North America, including Greenland, the West Indies, and Central America. Issued irregularly; 100 parts have ap-. peared. 1905- Prices of parts on request. Mycologia. Issued bimonthly; devoted to fungi; official organ of the Mycological Society of America. Now in its forty-fifth volume. 1909- $1.50 a single copy, $7.00 a volume. __ Addisonia. Colored plates of flowering plants, accompanied by popular descriptions; eight plates in each number, thirty-two in each volume. Now in its twenty-second volume. 1916- Free to members of the Garden; to others $10.00 a volume. Brittonia. Issued irregularly; devoted to taxonomy and related subjects. Official organ _ of the American Society of Plant Taxonomists. Now in its seventh volume. 1931- To members of the Society, $5.00 a volume; to others, $7.50. The Garden Journal of The New York Botanical Garden. (Replaces the Journal of The New York Botanical Garden.) Issued bimonthly; contains non-technical articles on bot- any and horticulture, illustrated. 1951- Free to members of the Garden; to others $2.00 a ~ volume. MEMOIRS OF THE NEW YORK BOTANICAL GARDEN VoL. 8, No. 3 ~ Vegetation of Nyasaland. Report on the Vernay Nyasaland Expedition ee ee ev ne L. J. Brass 161 < _ Plants Collected by the Vernay Nyasaland Expedition of 1946 J. P. M. Brenan and Collaborators 191 Issued 23 June 1953 PRINTED BY THE ScIENCE PREss LANCASTER, Pa. The Memoirs of The New York Botanical Garden are issued at irregular intervals in parts of various sizes. Approximately 500 pages will complete a vol- ume. The subscription price of volume 8 is $10. Number 3 may be purchased separately for $2.00. Authors of papers may obtain separate copies of their contributions, printed at the same time as the issue, at cost price. For further information address the editor: | H. W. RIicKETT The New York Botanical Garden New York 58, N. Y. Vol. 8 No. 3 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN June 23, 1953 VEGETATION OF NAYASALAND REPORT ON THE VERNAY NYASALAND EXPEDITION OF 1946 i. J. BRASS INTRODUCTION The Vernay Nyasaland Expedition of the American Museum of Natural His- tory, on which I represented the New York Botanical Garden, was sponsored and led by Mr. Arthur S. Vernay, a trustee of the Museum and a member of the Council of the Garden. Other members of the party were Dr. Harold E. Anthony, Chairman of the Department of Mammals, American Museum, and the late Capt. Guy C. Shortridge, Director of the Kaffrarian Museum, King William’s Town, South Africa. The history of Nyasaland may be said to have begun in 1859, when a party led by David Livingstone, the great explorer-missionary, ascended the Shire River by steam launch from the Zambezi and examined southern parts of the country. Missionaries and traders, following Livingstone, were soon in conflict with Arab slavers and their native allies, and in order to cope with this situation Britain, in 1889, proclaimed a protectorate over the Shire Highlands, in the southern part of the area now included in Nyasaland. Two years later, the name ‘Protectorate of British Central Africa was applied to all territories then under British influence north of the Zambezi, including, in addition to all of Nyasaland, parts of Tanganyika and Northern Rhodesia. In 1893 the title British Central Africa was confined officially to Nyasaland, and in 1907 the title was changed to Nyasaland Protectorate. Beginning in May and ending in October, our work in Nyasaland was car- ried out in the dry season of 1946 (Brass 1948). The principal zoological in- terest of the expedition was in mammals, but some atténtion was given to the collection of reptiles and amphibians, fishes, and insects. The plant collections of the expedition comprised 2004 numbers, including 235 of non-vascular crypto- gams. An average of 6.2 herbarium sets was collected in vascular plants. The taxonomic work on the flowering plants, ferns, and fern allies was under- taken by the Royal Botanic Gardens, Kew. This work is still in progress. Except for some few species which I could identify with reasonable certainty in the field, and the mosses, determined by Edwin B. Bartram, the plant names cited in this report rest on the authority of lists received from Kew. All but the following groups were determined by J. P. M. Brenan: ferns (F. Ballard); grasses (C. E. Hubbard); sedges (E. Nelmes); orchids (V. S. Summerhayes); Ranunculaceae, Menispermaceae, Nymphaeaceae, Capparidaceae, Violaceae, Kiggelaria, Pit- tosporaceae, Polygalaceae, Caryophyllaceae, Hypericaceae, Guttiferae, Diptero- Carpaceae, Geranium, Crassocephalum, Acanthaceae, Eriocaulaceae, Eriosema (E. Milne Redhead); Anacardiaceae (except Lannea and Sorindeia), Allophyllus, Lantana, Lippia, Rhynchosia (R. D. Meikle); Melianthaceae (B. Verdcourt); Brachystegia (A. C. Hoyle); Alepidea (H. Weimarck); Gesneriaceae (B. L. Burtt), Amaranthaceae (K. Suessenguth); Proteaceae (R. D. Meikle and J. P. M. Brenan); Lobelia (E. Wimmer); Agathisanthemum, Kohautia, Oldenlandia, Pavetta (C. FE. B. Bremekamp). - New species described, or being described, from the collections of the Vernay Expedition are indicated in the following pages by an asterisk. My serial collec- tion numbers are given for plants which are as yet undetermined. 161 % 162 MEMOIRS OF THE NEW YORK BOTANIC AL GARDEN [Vol. 8, No. 3 PHYSICAL FEATURES AND GEOLOGY From the prominence of its mountains and lakes, Nyasaland has been called the ‘‘Scotland of Africa’’ by an admirer of its scenic beauties. About 520 miles in length and from 10 to 130 miles in width, it has an area of approximately 37,890 square miles. To the northwest is Northern Rhodesia, to the north and northeast Tanganyika Territory, and bordering its southern half Mozambique or Portuguese East Africa. To quote from Dixey (1932): ‘*The dominant feature in the physiography of the country is the deep, trough-like depression, forming part of the Great Rift Valley, that traverses it from end to end; the greater part of this trough is occupied by Lake Nyasa and the remaining part by the Shire River. The country on either side of the trough is made up of high plateaux. For example, that lying west of the lake stands mainly between 3,300 and 4,400 feet above sea- level. Near Dedza it rises to about 5,000 feet and towards the northern end of the lake the Vipya, Nyika, Mafingi and Misuku uplands rise to altitudes of 6,000 to 7,500 or even 8,000 feet. South of the lake the Shire Highlands plateau is sur mounted by the Mlanje Mountains, rising to nearly 10,000 feet, and the Zomba up- lands to about 7,000 feet. ‘‘Lake Nyasa stands at an elevation of about 1,520 feet, and the adjacent lake plains rise to about 1,700 feet; the southern part of the rift, occupied by the Lower Shire, stands only at 200 to 300 feet above sea-level. “‘Owing to the manner in which it is traversed by the rift valley a large Proportion of the Protectorate is more or less of mountainous character; this is particularly the case along the sides of the rift valley, which are generally steep locally and even precipitous. Apart from the floor of the Upper and the Lower Shire Valley, almost the only country of fairly even surface comparable with the greater part of Southern and Northern Rhodesia is the Angoniland plateau, which indeed is merely an extension of the great Rhodesian plateau.”’ Lake Nyasa, the third largest lake in Africa, is 350 miles long, 20 to 50 miles wide, and near its northern end over 300 fathoms deep. Smaller lakes include Lake Chiuta and shallow, brackish Lake Chilwa on the eastern Portuguese border. Lake Chilwa has no outlet. With the exception of an area on the eastern Portuguese border which drains in part east to the Lujenda River and in part to Lake Chilwa, the whole of Nyasa- land is drained by streams that flow to the Rift and empty into either Lake Nyasa or its outlet the Shire River, and so, by the Shire, south to the Zambezi. The long- est river other than the Shire, the South Rukuru, has a length of about 170 miles. Many of the rivers run throughout the year in their upper course, but are intermit tent in their lower course through evaporation and absorption in the dry season. Only a few maintain a yearround flow throughout their entire length. Dixey (l.c.) states that the basement rocks of Nyasaland consist of a complex of schists and gneisses, probably Pre-Cambrian, invaded by various intrusives. Massive intrusions of syenite, well exemplified in Mlanje and Zomba mountains, constitute prominent orographical features. Isolated blocks of sediments and lavas of the Karroo System (Permain to Rhaetic) occur as relicts of denudation, preserved by down-faulting into the older rocks in the Lower Shire Valley, on Explanation of map on page 163 Map of Nyasaland showing routes and principal collecting localities of the Vernay Nyasaland Expedition. 1. Blantyre, Shire Highlands. 2. Zomba Plateau. 3. Likubula Gorge. 4. Luchenya Plateau, Mlanje Mountain. 5. Nchisi. 6. Nyika Plateau. 7. Kasungu. 8. Camp Chibotela, Chia area. 9. Cholo Mountain. 10. Chikwawa. 163 VEGETATION OF NYASALAND 1953] 36° 35° 34 33° 9° SCALE OF MILES o 10 20 30 40 0 10° 2 12° PORTUGUESE NYASA vViOow VLOW 9 = 3” ‘2 % > EAST AFRICA = 15° 16° I7° < K > a. x S) RRS 2 ° = sy z ae = 8 px AS NI oK 5.2 ap