Described from “Colorado, about the Georgetown and South Park Roads [Clear Creek and Park Cos.],” which Mead ([1876]) reiterated. The lectotype (MCZ) (Fig. 9) designated by Barnes and McDunnough (1916) is labeled “Colorado.” Scott (1998) suggested a type locality of “Tennessee pass, 3150 m (=10300 feet), Lake-Eagle Cos. Colorado.” Andrew D. Warren proposed Guanella Pass (Park/Clear Creek Cos.) (Pelham 2008), but maps do not show a trail through that area during the 1870s. Scott (2008b, 2008c) stated “evidently Kenosha Pass [Park Co.] or Guanella Pass [Clear Creek Co.], Colo.”

Affixed to the lectotype is a portion of Mead's original field envelope with the penciled date of “7–13.” This specimen was figured by Scudder (1874, Pl. 10, fig. 18) as “collected July 13, by T. L. Mead.” A damaged male paralectotype at MCZ with the same date as the lectotype, was dissected and figured by Scudder (1874, Pl. 11, figs. 10, 11). Other butterflies at MCZ that Mead collected during mid-July 1871 bear dates written in his hand in either pencil or ink using the same format. Scudder (1874, Pl. 10, figs. 16, 17) also figured as colorado two females collected by Mead, one of which is dated 28 August. All these specimens were possibly papered duplicates that Scudder obtained when he visited Edwards' home in December 1873. Most of Mead's specimens of Hesperia from Colorado were examined by Scudder, who identified some as Pamphila manitoba (=Hesperia comma manitoba (Scudder)) (Scudder 1874). This undoubtedly prompted Mead ([1876]) to apply this name to two of his specimens, which represent high-elevation phenotypes of the Front Range subspecies Hesperia colorado ochracea Lindsey.

On 13 July 1871 Mead was at Twin Lakes, where he walked “up to the head of the upper lake” to collect butterflies. All the material from Twin Lakes was mailed to Edwards from Oro City on 20 July 1871. The female specimen that Mead captured on 28 August was taken along the Denver & South Park Road northeast of Kenosha House in northern Park County (Table 1), probably at an elevation of about 2621 m (8600 ft.). Mead ([1876]) mentioned collecting colorado along this road “during the latter part of August.” Although the peak flight period of this butterfly is mid-August to early September, adults can emerge much earlier, especially during unusual conditions. Precipitation in Colorado during the spring of 1871 was below modern norms (Mock 1991). In fact, much of the United States was already suffering drought conditions by early spring. One of the worst natural disasters in history took place during October 1871 when fire storms swept across the Great Lakes region, causing the Great Chicago Fire (van den Dool 2012). Other captures by Mead in 1871 suggest that some other butterfly flight periods were advanced (see C. hagenii and A. helena, below).

To identify the lectotype (MCZ-ENT00015299) as the primary type of colorado, a red printed and handwritten label was affixed to its pin, which reads, “LECTOTYPE / Pamphila colorado / Scudder, 1874 / per Barnes & McDunnough (1916) / John V. Calhoun, 2013.”

Type locality. As defined by the lectotype, the type locality is restricted to Twin Lakes, along the northern shore of Upper Twin Lake (the western lake), Lake County, Colorado.

Discussion. Skinner and Williams (1924) observed that “Individuals from high altitudes are darker green below than the type.” Referring to Scudder's (1874) figure of the lectotype, Lindsey (1942) suspected that it “may very well be true colorado.” Indeed, the lectotype (Fig. 9) and paralectotype from Twin Lakes are more brightly colored than higher-elevation butterflies (Fig. 12), which have long been considered to represent nominotypical colorado. Many authors, including MacNeill (1975), Stanford (1981), and Tilden and Smith (1986), considered nominotypical colorado to denote high-elevation populations of small dark butterflies. The lectotype, however, is more evocative of butterflies that are currently recognized as the subspecies Hesperia colorado oroplata (Fig. 11), described as H. comma oroplata by Scott (1981) to define more brightly-colored lower-elevation populations “from the Arkansas River Valley south of Buena Vista to the Royal Gorge, south to the Sangre de Cristo Mts. of New Mexico, and the San Luis Valley of Colorado.” Unfortunately, Scott did not attempt to locate type specimens of Pamphila colorado (Scott 1998).

To establish a baseline of comparison from which to differentiate oroplata, and to prevent the use of the name colorado for populations at lower elevations, Scott (1998) suggested Tennessee Pass (3170 m/10,400 ft.) as the type locality of colorado, despite a lack of evidence that Mead had collected there. Without further explanation, Scott (2008b, 2008c) later revised the type locality to “evidently Kenosha Pass or Guanella Pass,” which are 3048 m (10,000 ft.) and 3557 m (11,670 ft.), respectively. Scott (1998) considered “true” colorado to occur “at high altitude in Colorado in mountains along the continental divide … from about 10,000 feet upward,” adding that “the darkest populations are at the highest altitudes of up to 12,400 feet or more.” As an example of this taxon, Scott (1998, fig. 15) figured a male from Loveland Pass, Summit Co., Colorado, 3779.5 m (12,400 ft.) in elevation. Although this approach reflected the popular concept of this taxon, it is not supported by the lectotype. Thanks to Mead's journal, we can now establish with certainty that the lectotype of colorado was collected at a lower elevation.

The primary character used by Scott (1981) to distinguish oroplata was the “lighter yellowish rather than greenish brown” coloration of the ventral hindwings (Scott 1981). Additional cited attributes were “slight differences in genitalia, antenna length, number of micropyle spines & developmental period.” This analysis, however, was based on the notion that colorado is restricted to higher elevations; populations that Scott (1975a, 1975b, 1986) considered to be genetically discrete. Scott's interpretation of colorado is inconsistent with the lectotype, which agrees with the original description of oroplata in that the ventral hindwings are “lighter yellowish,” not greenish brown like those from higher elevations (Figs. 9, 11, 12). Scudder's (1874) figure of the lectotype faithfully portrays its more brightly-colored ventral hindwings (Fig. 10). Moreover, Scott (2006b) remarked that colorado is “evidently biennial like other alpine butterflies.” This differs from oroplata and populations of “true” colorado, both of which are annually-brooded. Forewing lengths (base to apex) of the holotype of oroplata and the lectotype of colorado measure 15.0 and 14.4 mm, respectively.

The lectotype of colorado was collected at an elevation of roughly 2819 m (9250 ft.), within an area that Scott (1975) identified as a transition zone between oroplata and higher-elevation populations, which he and other authors attributed to nominotypical colorado. Fifty-six specimens that Scott collected within this area are deposited at MGCL (ex. W. W. McGuire colln.). They were captured 2.6 km (1.6 mi) north of the town of Granite, Chaffee County (2774 m/9100 ft.), and at Mt. Massive Trout Club, Lake County (2856 m/9400 ft.). These populations occur within 9.7 km (6 mi) of the type locality and represent typical colorado as defined by the lectotype. Any perceived differences between these populations and those within the range of oroplata are subtle at best. They form a smooth gradual cline, with populations at lower elevations producing slightly larger and tawnier adults with somewhat reduced dark maculation. Because the primary criteria that were originally used to differentiate oroplata also extend to the nominotypical subspecies, all these populations should be considered to represent H. c. colorado. As a result, the dusky biennial populations at higher elevations in Colorado require a new name (Warren & Calhoun 2015).

The female that Mead collected on 28 August, which was figured by Scudder (1874, Pl. 10, fig. 17), is applicable to H. c. ochracea. The second female figured as colorado by Scudder (1874, Pl. 10, fig. 16) actually represents Hesperia nevada (Scudder).

In addition to specimens from Nevada, California and Oregon, this species was described from “Colorado … on the mountains about the South park and in the Park itself [Park Co.],” which Mead ([1876]) reiterated. Barnes and McDunnough (1916) designated one of Mead's specimens as the lectotype and defined the type locality as “S. Park, Colo.” Affixed to the lectotype (MCZ) is a portion of Mead's original field envelope dated “June 12.” This specimen was figured by Scudder (1874, Pl. 10, fig. 1) as “collected June 12, by T. L. Mead.” Scudder (1874, Pl. 10, figs. 3, 4) also figured male and female paralectotypes, dated 17 June and 23 June, respectively. A male paralectotype at CMNH was collected by Mead on 16 June (it is from Mead's collection, but Scudder examined all of Mead's Hesperia prior to describing this species). Colorado specimens of nevada from Scudder's collection were possibly papered duplicates that he obtained from Edwards in 1873.

Another male paralectotype at MCZ bears a portion of Mead's field envelope dated “6/17.” Its claspers were figured by Scudder (1874, Pl. 11, figs. 3, 4), who stated that the specimen was “collected June 17, by T. L. Mead.” This specimen also bears a second clipping from Mead's envelope reading “Mr. B.” Mead's journal entry for that date indicates that Samuel M. Blair collected 70 butterflies. Blair (1842–1932), a miner and relative of the proprietor of Kenosha House, assisted Mead and W. H. Edwards in collecting and rearing butterflies from the area. “Mr. Blair is, I think, to be trusted to take good care of the insects and to collect more,” Mead reported in July 1871 (Brown & Brown 1996). Mead initially took ownership of Blair's butterflies, but Blair later accepted an offer from Edwards to collect specimens in exchange for parts of his butterfly book. Blair's specimens were thereafter considered to be Edwards' property, though Mead saw this as an unfair arrangement, especially since he physically managed Blair's activities in Colorado. Mead eventually became dissatisfied with Blair, complaining that he lacked “that extreme accuracy necessary for scientific observations” (Brown & Brown 1996).

On 12 June 1871, Mead collected several miles up Beaver Creek near Fairplay (Table 1). On 16 and 17 June, he walked south from Kenosha House along the Denver & South Park Road. On 23 June, he collected along the same road just southwest of Bradford Junction (Table 1). Mead mailed the lectotype specimen to Edwards from Bradford Junction on 20 June 1871.

To identify the lectotype specimen (MCZ-ENT00015299) as the primary type of nevada, a red printed and handwritten label was affixed to its pin, which reads, “LECTOTYPE / Pamphila nevada / Scudder 1874 / per Barnes & McDunnough (1916) / John V. Calhoun, 2013.”

Type locality. As defined by the lectotype, the type locality is restricted to along Beaver Creek, east/northeast of Fairplay, Park County, Colorado.

Described from a single male collected by Mead in “Colorado,” which Brown and Miller (1980) defined as “Georgetown, Clear Creek County, Colorado.” Scott (2008b) suggested “Georgetown or eastward toward Idaho Springs.” Although the holotype (CMNH) does not possess a date, it bears an identification label of W. H. Edwards reading “napa ♂ = dacotah.”

Soon after describing dacotah, Edwards (1872) questioned its recognition and concluded it was synonymous with the species he had previously described as Hesperia napa (=Ochlodes sylvanoides napa). Based on this treatment, Mead ([1876]) did not separately list any records of dacotah from Colorado, nor did Edwards ([1876]) list it among the species recorded by the Wheeler Survey. These taxa were not again recognized as different species until Barnes & McDunnough (1916) treated dacotah as a variety of P. mystic.

Regarding napa (which was then considered to include dacotah), Mead ([1876]) stated that he “found a few specimens in August near Georgetown.” One of Mead's specimens of napa at CMNH is dated “Aug 19” (see below). On 19 August, Mead was traveling from Idaho Springs — where he had arrived the previous day from Georgetown — to Apex Gulch (Table 1). Mead ([1876]) also mentioned the capture of another species at Idaho Springs on 19 August. These specimens were mailed to Edwards from Denver on 22 August 1871.

Type locality. Restricted to the vicinity of Idaho Springs, Clear Creek County, Colorado.

Described from specimens collected in 1864 by James Ridings at “Empire City [Clear Creek Co.], Colorado Territory.” One of Mead's specimens (CMNH) was designated as the neotype by Brown and Miller (1980), who implied the type locality to be “the vicinity of Georgetown, Clear Creek County, Colorado.” However, Miller and Brown (1981) and Pelham (2008, 2014) reiterated the type locality as “Empire City, Colorado,” and “Empire in Clear Creek County, Colorado,” respectively. Mead ([1876]) stated that he “found a few specimens in August near Georgetown.”

Affixed to one of Mead's specimens of napa from the Edwards collection (CMNH) is a portion of his original field envelope, which reads “Hesperia A / Aug 19.” In a letter to Edwards, Mead mentioned taking “Hesp A” on 19 August at Idaho Springs (Brown & Brown 1996). On 19 August, Mead was traveling from Idaho Springs to Apex Gulch (Table 1). These specimens were mailed to Edwards from Denver on 22 August 1871.

Type locality. Restricted to the vicinity of Idaho Springs, Clear Creek County, Colorado.

Described as a possible new species from “Colorado,” which Opler (1966) defined as “Turkey Creek Junction, Jefferson County, Colorado.” This type locality was proposed on the advice of F. Martin Brown, who examined the male syntype and concluded that it is “without doubt one sent to Henry Edwards by W. H. Edwards from the material collected by T. L. Mead in 1871.” Mead wrote of finding early stages of this species at Turkey Creek Junction in June 1871 (Edwards 1874). He later told H. Edwards, “only one of my [larval] specimens gave imago and it proved to be Ausonides …” (4.iv.1874, AMNH).

Two specimens (male and female) from H. Edwards' collection, acquired by AMNH in 1892, possess his handwritten labels identifying them as “types” of Coloradensis. The male syntype was designated as the lectotype by Johnson (1976). Henry Edwards received one of Mead's specimens of this butterfly from W. H. Edwards in early July 1871. In the accompanying letter, W. H. Edwards asked if “Anthocaris 1” was ausonides; in agreement, H. Edwards scrawled that name on the page (8–9.vii.1871, AMNH). It could be assumed that it was this specimen to which Edwards (1881) referred when he wrote, “I have long had in my possession a ♂ example of Anthocaris, which appears to be distinct from any known form, but I hesitated to describe it until more material might present itself.” However, Mead sent at least two additional Colorado specimens of ausonides to Edwards in 1872 (19.vi.1872, RC, AMNH). Edwards (1881) did not attribute the male type specimen to Mead and only casually mentioned that Mead's collection also contained this species. Henry Edwards evidently did not recall where he had obtained the male type specimen, which was received years before he described Coloradensis. He exchanged hundreds, if not thousands, of butterflies after 1871 and it is possible that the specimen he mentioned in the original description was received from another correspondent at a later date.

Edwards' collection catalog at AMNH does not list any specimens by the name Coloradensis. Entry no. 14 lists “Anthocaris ausonides” from California. Although Opler (1966) indicated that the lectotype of Coloradensis was labeled “14,” no such label or notation is currently affixed to the specimen. Possibly referring to the two Coloradensis that Mead sent in 1872, entry no. 4277 lists an “Anthocaris” collected by Mead during June in “Rocky Mts. Col.”

According to Edwards (1881), the female paralectotype was received from the physician James S. Bailey (1830–1883) of Albany, New York (Curtis 1884). Not only did Edwards correspond with Bailey, he visited Bailey's home and wrote his obituary (Edwards 1883). One of Bailey's sons, Theodore P. Bailey, also collected Lepidoptera. The elder Bailey provided Colorado Lepidoptera to several other contemporary entomologists, including Augustus R. Grote, who described new taxa from his moth specimens. Bailey did not personally travel to Colorado, but obtained “unprepared” specimens through another entomological correspondent who employed a resident of Colorado to collect butterflies and sugar for moths (30.vii.1878, FMNH). Unfortunately, no letters from Bailey were found among Edwards' correspondence at AMNH.

Type locality. Although the source of the lectotype is uncertain, this taxon has been associated with “Turkey Creek Junction” for nearly fifty years and the species is known to occur there. The type locality is more accurately defined as the vicinity of Bradford Junction (“Turkey Creek Junction,” within the present-day town of Conifer), Jefferson County, Colorado.

Described from “various localities, from So. Colorado to Montana and Dacotah (Bismarck).” Without explanation, Ferris (1971) attributed the type locality to “Durango, La Plata Co., Colorado.” Brown (1973) designated one of Mead's specimens as the lectotype (Figs. 13, 14) and defined the type locality as “northeastern corner of South Park, Park Co., Colorado.” Affixed to the lectotype (CMNH) is a portion of Mead's original field envelope dated “6/16.” It also includes a fragment of Mead's notation identifying the specimen as a male “[Phil]odice” (Fig. 15), which is consistent with Edwards' (1887b) comment that Mead originally considered these specimens to represent “a variety of Philodice.”

On 16 June 1871, Mead was lodging at Kenosha House, where he “found many Colias” during a walk of 4.8 km (3 mi) along the road “to the South Park” (Table 1). He apparently walked along the Denver & South Park Road (today's I-285 corridor) from Kenosha House over the summit of Kenosha Pass and down toward the South Park basin. Although Brown (1973) stated that the lectotype was from Edwards' collection, there is no such collection label on the specimen (Fig. 15). The presence of the envelope clipping infers that Holland acquired it from Mead and it was still papered. Mead mailed this specimen to Edwards from Bradford Junction 20 June 1871.

Type locality. As defined by the lectotype, the type locality is restricted to within 1.6 km (1 mi) south of the summit of Kenosha Pass, Park County, Colorado.

Discussion. Although most authors since Smith (1891) consigned hagenii to synonymy, Garth and Tilden (1963) and Pohl et al. (2010) resurrected this name to recognize subspecies of Colias philodice Godart. The original description of hagenii characterized a summer phenotype with broad black wing margins (Edwards [1884a]), though Edwards (1887a) eventually considered hagenii to be “seasonally dimorphic.” Referring to specimens that would later be described as hagenii, Mead reportedly observed, “if there could be such a thing as a yellow Eurytheme, this was it” (Edwards [1884a], [1884b], 1887a). Mead's concept of Colias eurytheme Boisduval was of the summer phenotype, as the spring form of this butterfly was then thought to represent a different species, Colias ariadne W. H. Edwards. In his original description, Edwards ([1884]) stated that hagenii was “between Philodice and Eurytheme … add orange and Hagenii would be indistinguishable from Eurytheme.”

There is still a great deal of disagreement regarding the status of eurytheme, eriphyle, and philodice. Although most lepidopterists now treat eriphyle as a subspecies of philodice, some consider it worthy of specific recognition. Likewise, the true nature of hagenii remains somewhat obscure and even Edwards' concept of this nominal taxon evolved over time. In his original description, Edwards (1884a) noted that some adults of hagenii have an “ochraceous tint.” He also mentioned that the coloration is sometimes “not yellow, but of a peculiar shade, a sort of buff-yellow (better chrome-yellow), a shade not seen in Philodice” (Edwards 1884b). He observed that males “frequently show more or less of [this coloration], and occasionally have a flush of orange.” During the 1880s, Edwards reared adults of hagenii from eggs that were received from Herman W. Nash of Pueblo, Colorado. Edwards (1887a) identified some of these reared adults as eurytheme, prompting him to consider hagenii as a yellow form of eurytheme (Edwards 1887b). He ultimately concluded that hagenii was synonymous with eriphyle, but still regarded the latter as a “tetramorphic form” of eurytheme (Edwards 1887b). The lectotype of hagenii is a slightly darker shade of yellow (Figs. 13, 14), but it is unclear if this coincides with Edwards' reference to an “ochraceous tint” or is simply due to the specimen's age and storage conditions. Many specimens from Holland's collection are discolored due to being stored in his cellar prior to their transfer to the Carnegie Museum in 1896–97 (Brown & Miller 1980).

Although the lectotype of hagenii (Figs. 13, 14) agrees with the original description, Fisher (2012) questioned its date (16 June) because he could not confirm the occurrence of the summer phenotype of C. p. eriphyle during mid-June in the Fairplay area. He argued that the specimen was probably not collected by Mead, but instead was reared from a batch of eggs that Edwards received from H. W. Nash. This is untenable, however, as the writing on the clipping affixed to the lectotype, including the penciled notation “Colo,” is in Mead's hand (Fig. 15). It is also difficult to disregard Mead's accompanying journal entry about finding “many Colias” on its date of capture (16 June), or the fact that he reportedly collected most of his specimens of philodice during June (Mead [1876]). Because Mead mailed specimens from Colorado on a regular basis, it is unlikely that he collected the specimen at a later date and incorrectly labeled its envelope.

Unusual weather conditions during 1871 could explain the presence of a summer phenotype of eriphyle near Kenosha Pass on 16 June. Other captures by Mead near Kenosha House suggest that the 1871 season was earlier than normal (see P. colorado, above, and A. helena, below). It is also possible that the specimen was not of local origin, but rather reached the area from lower elevations where the summer form appears earlier in the season.

Described from specimens “Taken in Colorado by Mr. Mead.” Brown (1934) suggested a type locality of “Mt. Elbert, Mosquito Pass, Mt. Lincoln and Hoosier Pass.” Brown (1965) designated a lectotype and restricted the type locality to “Mosquito Pass, Lake—Park Counties, Colorado.” Although the lectotype is undated, there are several other specimens in CMNH from Mead's collection that possess clippings from his field envelopes dated “7–20.” Brown (1965) suspected that on 20 July 1871 Mead was “en route between Twin Lakes, Lake County, and Fairplay, Park County,” noting that Mead was definitely on Mosquito Pass the following day. On 20 July, Mead explored in the vicinity of Oro City, within California Gulch, Lake County (Table 1). He walked “up the gulch and on the mountains,” but did not reach the higher elevations of Mosquito Pass, located 9.7 km (6 mi) northeast of Oro City. In a letter dated 20 July, which accompanied a box of specimens that had been collected since the last shipment, Mead attributed all the butterflies he captured that day to California Gulch (Brown & Brown 1996).

Two additional dated specimens of helena that Mead collected are deposited at PMNH. A male bears a label written by C. P. Whitney dated “July 3 1871” and a female bears a label written by F. G. Sanborn dated “7 10.” Mead was lodging at Kenosha House on 3 July when he mentioned collecting about 50 specimens of this species “further up and more in the mountains” (probably up Hoosier Creek valley west of the house) (Brown & Brown 1996). This is an early date for this species, suggesting that the 1871 season was advanced (this species does emerge in early July on occasion, as there are two fresh specimens at MGCL from Park County dated 3 July 1990 from an elevation of 3426 m/11,240 ft.). On 10 July, Mead climbed a peak south of Twin Lakes (Table 1) (Fig. 8), where he found “swarms” of this species at the summit (Brown & Brown 1996).

Type locality. Although Mead collected this species at multiple locations, the type locality is restricted to the vicinity of Oro City, California Gulch, Lake County, Colorado. This location is most suitable because of its proximity to the type locality previously proposed by Brown (1965) and its association with several specimens at CMNH where the lectotype is deposited.

Described in part from a male specimen taken by Mead at “Turkey Creek Junction [Jefferson Co.], in Colorado,” which Brown (1934) restricted to “Turkey Creek Junction, Colorado … probably the junction of South Turkey Creek with Turkey Creek, a few miles up in the foothills.” Mead's male specimen (CMNH) was designated by Brown (1965) as the lectotype. According to Mead ([1876]), it was collected on 6 June 1871. On that date, Mead was at Bradford Junction (“Turkey Creek Junction”) where he “took a lunch along & started down the road,” walking 12.9 km (8 mi), then returned by the same route (Table 1). He stated in a letter (Brown & Brown 1996) that he followed Turkey Creek, apparently walking along the winding Denver & South Park Road (today's South Turkey Creek Road), which paralleled South Turkey Creek northeast of Bradford Junction (Fig. 7). The lectotype specimen was mailed to Edwards from Bradford Junction on 20 June 1871.

Type locality. Restricted to within 9.7 km (6 mi) northeast of Bradford Junction (“Turkey Creek Junction,” within the present-day town of Conifer), along South Turkey Creek, Jefferson County, Colorado.

Described from “about 20 specimens taken in Colorado, in August 1871, by Mr. Mead.” Brown (1934) defined the type locality as “near Berthoud Pass, with little doubt on the southern slope [Clear Creek Co.].” The lectotype (CMNH), designated by Brown (1967), bears a portion of Mead's original field envelope dated “8/16.” On 16 August 1871 Mead was approaching Berthoud Pass from the north along the Empire & Middle Park Wagon Road (Table 1), when he stopped and collected 15 specimens of a new Grapta. Mead recorded in his journal that these specimens were collected “a little before reaching the pass.” Two paralectotypes, which Mead collected on the same date, are deposited at CMNH (ex. Mead colln.). These specimens were mailed to Edwards from Denver on 22 August 1871.

Type locality. As defined by the lectotype, the type locality is restricted to the northern slope of Berthoud Pass, Grand County, Colorado.

Described from “Colorado,” which Brown (1934) defined as “Twin Lakes, California Gulch, Fairplay and probably Kenosha. Probably not at Turkey Creek Junction.” Miller and Brown (1981) suggested “South Park, Park Co., Colorado.” Scott et al. (2006) suggested “South Park, probably near Fairplay, Park Co., Colo.” Fisher (2006) proposed “the vicinity of Fairplay or certainly between Fairplay and ‘Kenosha House’ within South Park [Park Co.].” Although Mead ([1876]) clearly intended to credit the name Melitaea eurytion to W. H. Edwards, modern rules of nomenclature (ICZN 1999, Art. 50.1) dictate that Mead be recognized as the author. For similar reasons, Barnes and Benjamin (1926) also attributed authorship to Mead.

Among the specimens of this taxon at CMNH is a somewhat rubbed male (Figs. 16, 17), which was figured as a “type” by Holland (1931, Pl. 57, fig. 15) and as “typical” by Scott et al. (2006, Pl. 5, top row). It bears a portion of Mead's original field envelope, reading “Melitaea 1” with the date “7/4” written in the same purple ink that Mead used for his journal entry on that date (Fig. 18). On 4 July 1871 Mead was lodging at Kenosha House (Table 1). This specimen was mailed to Edwards from Fairplay on 7 July 1871.

An undated female at CMNH (Figs. 19, 20) was also figured as a “type” by Holland (1931, Pl. 57, fig. 16) and as “typical” by Scott et al. (2006, Pl. 5, second row). This specimen bears a small label in Mead's hand identifying it as “Melitaea Eurytion Edw” (Fig. 21), suggesting it was mounted prior to being purchased by W. J. Holland. It also bears a larger typed label, probably prepared by Holland, which refers to the specimen as a “typr” (sic; type) with the notation “= nubigena fide Strkr,” which likely refers to Strecker (1878), who wrote “Eurytion = Nubigena.” Holland's (1931) figure of this specimen is too reddish, causing it to resemble the lower-elevation subspecies Euphydryas anicia capella (W. Barnes). New images from CMNH (Figs. 19, 20) reveal that it is actually quite dingy, probably due to its early storage conditions. Many specimens from Holland's collection were discolored due to being stored in his cellar prior to their transfer to the Carnegie Museum in 1896–97 (Brown & Miller 1980).

Scott et al. (2006) doubted the occurrence of this butterfly at Kenosha Pass, but adults were photographed there in early July 2008 by Thomas Bentley (Harp 2009, Warren et al. 2012), who indicated (pers. comm.) that they were “quite common.” On 9 July 2014, Andrew D. Warren found this species to be abundant within the meadows along Hoosier Creek directly opposite the former site of Kenosha House, at elevations from 2970 and 3000 m (9744 and 9843 ft.). This species was evidently also common in that area in 1871, as at least twelve other specimens at CMNH were captured by Mead in the vicinity of Kenosha House between 14 June and 2 July. Two additional females, which Mead collected in the same area, are deposited at PMNH (YPM ENT429625 and YPM ENT799979) (Figs. 22, 23, 25, 26). These specimens are from the collection of C. P. Whitney, who received them from Mead in early 1872 (6.iii.1872, RC). Both bear Whitney's labels that attribute them to Mead (Figs. 24, 27). In its original condition, the discolored female at CMNH, identified by Mead as M. eurytion, was phenotypically analogous to the female at PMNH dated 4 July (Figs. 25, 26). Mead's specimens from the vicinity of Kenosha House were likely collected at elevations between 2941 and 3200 m (9650 and 10,500 ft.). Also included among Mead's material at CMNH are three specimens from near Fairplay (Park Co.) and one from Twin Lakes (Lake Co.). At the request of John E. Rawlins, I provided locality information for all of Mead's dated specimens of E. anicia from Colorado at CMNH.

Discussion. Reflecting his frustration with identifying specimens, Mead declared, “Melitaea is I think about the most difficult genus we have …” (15.xii.1872, RC). His idea of M. eurytion applied to a particular color variation of the butterfly currently represented by this name and likely involved a second taxon. A review of the history of Melitaea eurytion is helpful in understanding Mead's concept of this butterfly.

On 9 July 1871, William H. Edwards (WHE) sent a batch of Mead's Colorado specimens to Henry Edwards (HE) for his opinion about their identities. In the associated letter, WHE asked if the species identified as “Melitaea 1” was “anicia or what,” to which HE scrawled “not anicia” (AMNH). WHE informed Mead on 25 August 1871 (MGCL) that HE had determined the species as Melitaea nubigena Behr (=Euphydryas editha nubigena; a California taxon). In the collection catalog of HE (AMNH) is an entry for “Melitaea nubigena” from “Rocky Mts, Colorado” from “T. Mead,” but these may refer to four other specimens, identified as “nubigena”, which Mead sent during 1872 (18.i.1872, RC; 19.vi.1872, AMNH). The two female specimens at PMNH from the Whitney collection are labeled as Melitaea nubigena, as originally identified by Mead on the advice of HE (Figs. 24, 27). Still confused, WHE wrote on 21 March 1873, “There are some remarkable forms among Mead's Colo [Colorado] Melitaeas that I had labelled as Nubigena & I think there are 2 species mix'd“ (AMNH). WHE was clearly flummoxed by the array of melitaeine phenotypes found by Mead and even proposed to call several specimens Melitaea idas, but he never published this name.

In late 1873, WHE proposed the name Melitaea eurytion (14.xii.1873, AMNH), but it was Mead ([1876]) who actually published it. Crediting the name to WHE, Mead wrote only two sentences about eurytion: “This species is found associated with Nubigena in Colorado, but is much rarer, and does not seem to range to quite so great elevations. The most obvious point of distinction from Nubigena is that the yellow spots of the latter are largely obscured in Eurytion by fulvous.” Despite Mead's publication of the name, Melitaea eurytion was subsequently abandoned. Strecker (1878) was the first to argue that Mead's ([1876]) comments were not sufficient to serve as an original description. Even WHE disregarded eurytion and instead maintained that all these butterflies from Colorado were either Melitaea anicia E. Doubleday or M. nubigena, and he did not list eurytion as a synonym of either (Edwards [1876], 1877, [1885]). Melitaea eurytion was not included in any checklists until Barnes and Benjamin (1926) considered it to be a synonym of anicia. Holland (1931) resurrected the name to define specimens “from comparatively low altitudes” in Colorado. He figured two of Mead's specimens as “types,” thus establishing the popular concept of this taxon. Brown et al. (1955a) also regarded eurytion as a full species. Miller and Brown (1981) considered eurytion to be a subspecies of anicia, yet acknowledged the name's tenuous nomenclatural availability. This treatment is generally accepted today.

Mead ([1876]) considered the butterfly that he called Melitaea nubigena to be “quite common throughout the mountain district of Colorado.” He described several larvae that he believed to be this species, one of which suspended itself on 19 June and pupated the following day. He mentioned these same larvae in a letter to W. H. Edwards on 22 June, stating that he had found them in South Park (Brown & Brown 1996). From 8 to 20 June Mead explored around Fairplay and Kenosha House, where he collected many adults of this butterfly. His concept of nubigena was clearly synonymous with the taxon now known as E. a. eurytion and probably included populations from above timberline, which are currently recognized as the subspecies E. a brucei (W. H. Edwards). Mead ([1876]) also listed Melitaea anicia from Colorado, but stated, “I did not meet with it in the Territory.” Like WHE, Mead apparently associated the darkest high-elevation phenotypes (i.e. extreme brucei) with nominotypical anicia. Lastly, Mead ([1876]) tentatively identified a single specimen from Bradford Junction as “Melitaea chalcedon” (=Euphydryas chalcedona E. Doubleday), but this was presumably based on an unusually dusky or aberrant individual.

In late 1873, WHE informed Mead of his concept of eurytion. Mead responded, “I found I could separate what I take to be Eurytion quite readily from the Nubigenas by the general appearance but in the details of markings I was unable to satisfactorily find the precise manner of difference except indeed in the greater proportion of yellow in Nubigena.” By introducing the name Melitaea eurytion, Mead ([1876]) clearly attempted to segregate the more fulvous butterflies within montane populations of “nubigena” (=anicia) in Colorado. The undated female specimen in CMNH, which Mead personally identified as M. eurytion (Figs. 19–21), is consistent with this concept in having its pale spots “largely obscured with fulvous.” Likewise, the pale spots of the analogous female from PMNH, collected on 4 July (Figs. 26, 27), are broadly infused with fulvous scales. Brown (1966b) supposed that Mead used the name eurytion to denote the “red form” of E. a. brucei, but Mead's named female does not agree with brucei, whose alpine populations contradict his account that eurytion “does not seem to range to quite so great elevations.” Adults currently found near the former site of Kenosha House are extremely variable, with some showing characteristics of E. a. capella and a few resemble E. a. brucei. Populations in the vicinity of Kenosha Pass embody the popular notion that eurytion is a highly variable taxon (see Ferris 1981).

Mead's concept of eurytion probably extended to lower-elevation populations of E. a. capella. Although capella is variable, its yellow (or yellowish-white) pattern elements are typically fulvous, corresponding with Mead's original description of eurytion. On 21 June, while exploring along the road just south of Bradford Junction (“Turkey Creek Junction”), Mead mentioned finding “another species of Melitaea of the same size as No. 1, but without any yellow spots” (Brown & Brown 1996). At CMNH is a dated male capella which Mead collected along the same road two days later. In addition, a male capella was figured by Wright (1905, Pl. 19, fig. 165) with the statement “no data, from Colorado; T. L. Mead.” This specimen (Figs. 28, 29) is deposited among the Wright “plesiotypes” at CAS and was listed by Tilden (1975). Unfortunately, it bears only one small label reading “165,” which corresponds to the species/figure number used by Wright (1905) (Fig. 30). This specimen was most likely also captured in the vicinity of Bradford Junction. Because Mead ([1876]) did not separately mention fulvous lower-elevation populations (i.e. capella), it is reasonable to assume that he considered all ruddy phenotypes to represent eurytion.

Lectotype designation. To preserve nomenclatural stability and fix the type locality, I hereby designate the male syntype at CMNH (Figs. 16, 17) as the lectotype of Melitaea eurytion, Mead, [1876] in accordance with ICZN (1999, Art. 74.7). This specimen was selected because 1) it is from Mead's personal collection, 2) it is dated, 3) it is from an extant population where Mead collected many of these butterflies, 4) its illustration as a “type” by Holland (1931) helped forge the modern concept of eurytion, and 5) Scott et al. (2006) referred to this specimen as “typical” and nearly designated it as the lectotype. The specimen bears 1) a white handwritten rectangular label [Eurytion /♂] in the hand of W. J. Holland, 2) a small rectangular printed and handwritten label [Butterfly Book / Pl. 57 Fig. 15], 3) a small rectangular printed label [Collection / T. L. Mead], and 4) a portion of T. L. Mead's original field envelope, reading “Melitaea 1 / 7/4” in faded purple ink (Fig. 18). A red printed label [LECTOTYPE / Melitaea eurytion / Mead [1876] / Designated by / John V. Calhoun, 2014] (Fig. 18) has been affixed to the specimen. All other specimens of eurytion collected by Mead at CMNH, as well as those at PMNH, are paralectotypes. The specimens sent to Henry Edwards were not located by staff at AMNH. Additional paralectotypes of eurytion likely exist, as Mead's letter copybooks (RC) indicate that he also sent specimens of “nubigena” to J. Behrens, H. K. Morrison and others. Mead collected 297 specimens of “Melitaea and Phyciodes” in Colorado (Mead 1877).

Euphydryas anicia carolae T. Emmel & L. N. Harris is treated by most authors (e.g. Fisher 2006, Scott et al. 2006, Pelham 2008, 2014) as synonymous with M. eurytion. The type locality of carolae is 2.1 km (1.3 mi) south of Fairplay at an elevation of 3005 m (9860 ft.) (Emmel & Harris 1998). This location is about 32 km (20 mi) southwest of Kenosha House, where Mead collected the lectotype of eurytion. Fisher (2006) suggested that Emmel and Harris (1998) disregarded eurytion in the tradition of prior authors, but Thomas C. Emmel (pers. comm.) confirmed that he recognizes eurytion as a valid subspecies and carolae was described solely to differentiate populations of brightly colored anicia that occur in dry-meadows within open valley floors of northern South Park. However, Scott et al. (2006) observed that most adults from the carolae type locality “look like the eurytion types (some quite like the types).” Moreover, some of the adults found around Kenosha Pass are very similar to carolae (Andrew D. Warren, pers. comm.). The various phenotypes of anicia in Colorado need further study to better understand the extreme variability between, and within, populations.

Type locality. As defined by the lectotype, the type locality is restricted to the vicinity of Kenosha House, 2.16 km (1.34 mi) north of the summit of Kenosha Pass, Park County, Colorado.

Described from “Colorado and Arizona,” Brown (1966) designated a lectotype from one of Mead's specimens and defined the type locality as “Turkey Creek Junction, Jefferson County, Colorado.” The lectotype (CMNH) bears Mead's small identification label dated “June 26,” indicating that it was probably mounted prior to being purchased by Holland. On 26 June 1871 Mead collected butterflies along the Mt. Vernon Wagon Road and “up a small brook” north of Bradford Junction (Table 1). Specimens collected that day were mailed to Edwards from Fairplay on 7 July 1871.

Type locality. As defined by the lectotype, the type locality is restricted to within several kilometers north of Bradford Junction (“Turkey Creek Junction,” within the present-day town of Conifer), Jefferson County, Colorado.

Edwards first proposed the name calydon in 1873 (14.xii.1873, AMNH), but never published it. Citing Edwards, Mead ([1876]) listed M. calydon without description in association with specimens that he collected at “Turkey Creek Junction.” Holland (1931) resurrected the name calydon to define Colorado specimens that resembled Melitaea (=Chlosyne) palla Boisduval. He figured a pair of “types” and, following Mead ([1876]), suggested a type locality of “Turkey Creek Junction, Colorado.” Mead ([1876]) reported that he collected specimens at that locality during 20–30 June 1871. Although he arrived at that locality on 20 June, he actually departed on 28 June (Table 1). Mead mailed the lectotype specimen to Edwards from Fairplay on 7 July 1871.

Type locality. The type locality is more accurately defined as the vicinity of Bradford Junction (“Turkey Creek Junction,” within the present-day town of Conifer), Jefferson County, Colorado.

Described from specimens “Taken in Colorado by Mr. Mead.” Brown (1966) designated a lectotype and defined the type locality as “Fairplay, Park County, Colorado.” Affixed to the lectotype (CMNH) is a portion of Mead's original field envelope, dated “7–4,” written in the same purple ink as Mead's journal entry for that date. On 4 July 1871 Mead was lodging at Kenosha House (Table 1). He mailed the lectotype specimen to Edwards from Fairplay on 7 July 1871.

Type locality. As defined by the lectotype, the type locality is restricted to the vicinity of Kenosha House, about 2.16 km (1.34 mi) north of the summit of Kenosha Pass, Park County, Colorado.

Described from “several specimens taken in Colorado by Mr. Mead.” Brown (1966) designated a lectotype and defined the type locality as “Denver, Denver County, Colorado.” The lectotype (CMNH) bears a portion of Mead's original field envelope with the date “6/3.” On 3 June 1871 Mead was lodging in Denver, where he “walked out on the prairie and found plenty butterflies” (Table 1). He followed Cherry Creek, a tributary of the South Platte River that runs southeast from the center of the city. The lectotype specimen was mailed to Edwards from Denver the same day on which it was collected.

Type locality. Restricted to Denver, along Cherry Creek, Denver County, Colorado.

Described from “Colorado; from several specimens taken by Mr. Mead.” Brown (1934) suggested the type locality of “Fairplay [Park Co.].” Brown (1964) designated a lectotype (CMNH), which bears a portion of Mead's original field envelope dated “7–6.” On 6 July 1871 Mead collected “up” Beaver Creek near Fairplay (Table 1). This specimen was mailed to Edwards from Fairplay the following day.

Type locality. As defined by the lectotype, the type locality is restricted to along Beaver Creek, east/northeast of Fairplay, Park County, Colorado.

Described from specimens “taken in Colorado by Mr. Mead.” Brown (1934) suggested a type locality of “Twin Lakes,” as well as “Mt. Lincoln, South and Middle Park.” Brown (1969) designated a lectotype and defined the type locality as “vicinity of Twin Lakes, Lake Co., Colorado.” The lectotype (CMNH) bears a portion of Mead's original field envelope, which is dated “7–13.” On 13 July 1871 Mead was indeed at Twin Lakes (Table 1). He walked “up to the head of the upper lake,” where he found “lots of Chrysophanus” He collected at least 40 males and two females in a “grassy spot” (Mead [1876], Brown & Brown 1996). Additional specimens of sirius which Mead collected during his stay at Twin Lakes are deposited at CMNH (ex. Mead colln.), MCZ (ex. Scudder colln. via Morrison), and PMNH (ex. Whitney colln.). These specimens were mailed to Edwards from Oro City on 20 July 1871.

Type locality. As defined by the lectotype, the type locality is restricted to Twin Lakes, along the northern shore of Upper Twin Lake (the western lake), Lake County, Colorado.

Described from three specimens “Taken in Colorado by Mr. Mead.” Brown (1934) cited Mead ([1876]), who stated that specimens were taken “on the South Park road four miles from the park.” Shields (1966) defined this as “ca. 1 mi. E Kenosha Pass summit, U.S. Highway 285,” which Brown (1970a) cited as “about 1 mile east of Kenosha Pass, Park County, Colorado.” Although the lectotype is not dated, Mead ([1876]) reported that he caught all of his specimens at the same locality on 17 June 1871. On this date, he was collecting along the Denver & South Park Road near Kenosha House (Table 1). His comment about taking the specimens “four miles from the park” denoted road miles from the South Park basin. Although the suggested type locality of Shields (1966) is essentially correct, the old wagon road (and existing U.S. Hwy 285) actually crossed the pass in a more north/south direction, not east/west. One mile east of the pass would place the collection site far off the road among the nearby peaks. Mead's specimens of ninus were mailed to Edwards from Bradford Junction on 20 June 1871.

Type locality. Restricted to the vicinity of Kenosha House, about 2.16 km (1.34 mi) north of the summit of Kenosha Pass, Park County, Colorado.

Described from specimens “Taken in Colorado by Mr. Mead.” Based on Mead ([1876]), Brown (1934, 1970b) defined the type locality as “Turkey Creek, Jefferson Co., Colorado.” Mead collected only three specimens during the last week in June, two of which are deposited at CMNH (ex. Edwards colln.). Neither is dated, but Mead spent 20–28 June at “Turkey Creek Junction” (Table 1), where on several occasions he walked along South Turkey Creek (Turkey Creek). His specimens from this period were mailed to Edwards from Fairplay on 7 July 1871.

Type locality. Based on the location of the settlement and Mead's documented activities, the type locality is defined as along South Turkey Creek, within 9.7 km (6 mi) northeast of Bradford Junction (“Turkey Creek Junction,” within the present-day town of Conifer), Jefferson County, Colorado.

Described from “many specimens taken in Colorado by Mr. Mead, in the summer of 1871.” Although Edwards (1873) mentioned that he had also received this species from Nevada and Arizona, it was implied that the description was based solely upon Mead's specimens from Colorado. Brown (1934) suggested a type locality of “Fairplay [Park Co.],” which Nabokov (1949) expanded to “Park Co., between Fairplay and Californian Gulch, on the way to Mosquito Pass, 9,500–13,188 ft. alt.” Brown (1970b) designated a lectotype (Fig. 31) and defined the type locality as “vicinity of Twin Lakes, Lake Co., Colorado … probably taken on La Plata Peak,” which Miller and Brown (1981) reinterpreted as “La Plata Peak, vic. Twin Lakes, Lake Co., Colorado.” Scott (2006a) suggested “Tinytown [Jefferson Co.].” Andrew D. Warren agreed with Brown (1934) that Fairplay is a more likely type locality (Pelham 2008).

The lectotype (CMNH) designated by Brown (1970b) bears a portion of Mead's original field envelope dated “7–11.” Mead was at Twin Lakes on 11 July 1871, but he did not climb any nearby peaks (Table 1). He was stiff after ascending a peak the previous day and instead collected in the vicinity of the lodging house (east of the lakes) and walked 4 km (2.5 mi) “above” the house, presumably northwestward along the road (now State Hwy 82). Mead remarked that this species was “very plentiful” around Twin Lakes, where it was “abundant by the first week in July” (Mead [1876], Brown & Brown 1996). In addition to the lectotype, at least two males and five females of melissa which Mead collected around Twin Lakes are deposited at CMNH (ex. Mead colln.). These specimens, including the lectotype, were mailed to Edwards from Oro City on 20 July 1871.

Type locality. As defined by the lectotype, the type locality is restricted to Twin Lakes, along the eastern side of Lower Twin Lake (the eastern lake), Lake County, Colorado.

Discussion. The type locality of Lycaena melissa (Fig. 31) is located about 16 km (10 mi) east of the type locality of Lycaeides melissa pseudosamuelis Nabokov (Fig. 32). The proximity of the type localities and similarity of the primary types prompted Scott (2006a) to speculate on the reasons why a drab high-elevation phenotype was chosen as the lectotype of melissa, even suggesting that F. Martin Brown was “peeved at Nabokov” and deliberately intended to force pseudosamuelis into synonymy. Although Brown (1950a, 1950b) disagreed with Nabokov's statistical methods, I have found no evidence that he had any such deliberate intentions regarding melissa. Appreciating Brown's motivation requires a review of the complex conceptual histories of the nominal taxa Lycaena melissa and Lycaeides melissa pseudosamuelis.

On 4 August 1871, W. H. Edwards informed Mead that three of his Lycaena species appeared to be new, but cautioned that he was not yet certain (RC). In early 1873, Edwards was unsure of the identity of Mead's “Lycaenas,” particularly one “like Scudderii, Edw., that has passed for Anna”(16.i.1873, AMNH). Edwards was referring to Lycaena scudderii (=Plebejus idas scudderii) and Lycaena anna (=Plebejus anna), both of which he had previously described. Seeking another opinion, he sent several specimens to Henry Edwards, referring to them as “all of Mead's collecting in Colorado.” After comparing his type of anna with Mead's specimens, W. H. Edwards wrote, “certainly the two seem distinct and I think the Coloradian is undescribed” (23.i.1873, AMNH). Within two months he had chosen a name: “The Lycaena like Anna I call Melissa” (9.iii.1873, AMNH). One of the male specimens that W. H. Edwards sent to H. Edwards eventually made its way back to Mead. It is deposited at CMNH (Brown 1970b) and was figured by Holland (1898, 1931; Pl. 31, fig. 25). Another specimen, a female, is deposited at AMNH and labeled “4880” in H. Edwards' hand. It corresponds to an entry in his collection catalog (AMNH) for an unidentified species of “Lycaena,” which was collected by Mead during June in “Rocky Mts. Colorado.”

In his original description of Lycaena melissa, Edwards (1873) defined a butterfly with a “complete sub-marginal row of large orange spots” on the ventral hindwings, as well as a row of orange spots on the ventral forewings. He characterized the female as having a “complete orange band” on the dorsal forewings and hindwings. However, Edwards typically chose one specimen of each gender from which to derive his original descriptions (Brown 1965). To define melissa, he presumably selected a pair that significantly differed from his concepts of scudderii and anna (Edwards 1873). Lower-elevation populations of melissa tend to be more brightly colored. Mead's specimens of melissa at CMNH suggest that he collected more from high elevations than from low, supporting the idea that Edwards cherry-picked the specimens he described.

Edwards' correspondence reveals that he intended to illustrate melissa with his description, but did not do so. Instead, he figured dorsal and ventral aspects of the male and female in Mead ([1876], Pl. 36, figs. 5–8) (Fig. 34). Edwards prepared the plates for that publication and undoubtedly selected the figured specimens (Calhoun 2013b). Because Edwards had exchanged much of his share of Mead's 1871 material by that time, the melissa specimens he figured were probably not those characterized in the original description. None of the melissa at CMNH were identified by Edwards as type specimens (Brown 1970b).

Barnes and McDunnough (1916) examined specimens at CMNH, including “presumable types from Colorado” and defined typical melissa as having “heavy red submarginal markings on the underside.” They considered specimens in their possession from Colorado, Utah, Arizona, and New Mexico as “typical.” Holland (1931, Plate 66, fig. 17) figured a brightly colored female from Colorado and identified it as a “type.” Today, nominotypical melissa is generally recognized as a low-elevation taxon with well-developed ventral orange spots that occurs over a large portion of western North America (Downey 1975, Fisher 1981, 2009, Scott 1986, Lane & Weller 1994, Layberry et al. 1998, Gompert et al. 2008) (Fig. 35). Fisher (1981) considered this taxon to be most common in the Rocky Mountains at elevations of 1525–2135 m (5000–7000 ft.). Despite our modern interpretation of nominotypical melissa, this taxon was treated differently during the four decades between the publication of Holland (1931) and Downey (1975).

Around 1932, F. Martin Brown began researching Mead's itinerary in an effort to clarify the type localities of taxa attributed merely to “Colorado” by W. H. Edwards. Regarding melissa, Brown (1934) stated, “Mead gives Fairplay as the probable locality for the 1871 series.” This assertion was derived from Mead's ([1876]) remark that specimens “were brought in by the expedition of 1871, probably from the vicinity of Fairplay.” Brown, however, mistakenly assumed that Mead was a member of a larger survey party and likely interpreted Mead's comment to mean that another party member had collected the type series, but the locality was unknown. Rather, Mead's uncertainty about the locality implies that he was referring to a different expedition than his own. In several instances Mead ([1876]) cited the “expedition of 1871” in reference to other exploring parties. Mead was probably referring to the Allen Expedition, who was in the vicinity of Fairplay in late July 1871 (Brown et al. 1956). It is also possible that Mead intended to cite the Wheeler Expedition of 1873, whose party collected in South Park, including Fairplay, during July and August (Yarrow [1876b]). Mead also listed melissa from Denver, and “In the South Park and about Twin Lakes.” Based on what we now know about his itinerary, he was referring to his own specimens from those areas. Nonetheless, Brown's concept of nominotypical melissa was forever biased by Mead's reference to Fairplay, which is roughly 3048 m (10,000 ft.) in elevation.

While Brown (1934) began to shift the concept of nominotypical melissa, it was Nabokov (1949) who formally advocated the change. Among the melissa specimens that Nabokov examined at MCZ was a female collected by Mead on 8 July 1871. Nabokov stated that “Mead's label” read “Divide between the Arkansas and S. Platte Valleys.” This specimen supported Brown's (1934) assertion that the type material was from the vicinity of Fairplay. In addition, Brown (1934) placed Mead on 8 July 1871 “on the divide between Fairplay and California Gulch” (i.e. Mosquito Pass), which corresponds with the data on the specimen. Nabokov therefore considered the type locality of melissa to be “Park Co., between Fairplay and Californian Gulch, on the way to Mosquito Pass, 9,500–13,188 ft. alt.” Accordingly, Nabokov (1949) regarded Mead's female specimen as a topotype and figured it on his Plates 6 and 9. I examined images of this specimen and found that its label was not written by Mead, but by Herbert K. Morrison, who received the specimen from Mead on 20 February 1873 (Mead referred to it in his letter to Morrison as “L. anna?”) (RC). Morrison probably sent this specimen to S. H. Scudder, whose collection is deposited at MCZ. Morrison's label, based on data provided by Mead, reads, “Divide between the / Arkansas & Platte / valleys July. 8.1871.” In an effort to reconcile the locality label with Brown's (1934) allusion to Fairplay, Nabokov affixed a label to this specimen which reads, “vic. Fairplay, / Mosquito Pass / (see Brown 1934 J. N.Y. / Ent. Soc.) leg T. L. Mead.” It is no secret that Nabokov greatly admired Brown's studies of Mead (Boyd & Pyle 2000).

Influenced by Brown's assumptions and encouraged by Mead's specimen, Nabokov (1949) concluded, “typical melissa is not the low level (sage belt, oak brush, alfalfa, prairie, etc.) form or forms, with richly ornamented underside and female upperside, but an altitudinal, comparatively drab race, little known to collectors.” To conserve the “general concept of melissa,” he considered the “richly ornamented” low-elevation populations as an extreme form. A male melissa, collected by Mead on 22 July 1871 between Fairplay and Montgomery City, Park Co., Colorado (PMNH, ex. C. P. Whitney colln.) (Fig. 33), is consistent with nominotypical melissa as interpreted by Brown (1934) and Nabokov (1949).

Nabokov (1944) was the first to mention “a curious Colorado form” of melissa with a reduced ventral orange band. He later described this form as the subspecies L. m. pseudosamuelis to “delimit the geographically adjacent type form of melissa melissa on the negative side of its pattern” (Nabokov 1949). He admitted that pseudosamuelis (Fig. 32) was the “weakest of melissa races in Colorado,” with the “Fairplay typical form coming next, and the S. E. Colo. form bringing ssp. melissa to its maximum expansion in Colorado” (Figs. 33, 35). In essence, Nabokov described pseudosamuelis to narrow his definition of the nominotypical subspecies. Nabokov (1955) later characterized a population of melissa from above timberline in the Sierra Madre Range of Wyoming as a “colony of typical (alpine) L. melissa melissa as described by Edwards,” thus confirming his belief that typical melissa is a high-elevation taxon separate from pseudosamuelis, which he limited to Pitkin and Lake Counties of Colorado.

Ten of the 17 type specimens of pseudosamuelis, including the holotype (Fig. 32), were collected by the Kansas lepidopterist John R. (J. E. Rice ) Turner (1910–2000) along US Hwy 82 in the vicinity of the Red Mountain Inn, Lake Co., Colorado. The former Red Mountain Inn was a late 19th century stage stop along the Twin Lakes toll road, 9.7 km (6 mi) west of the town of Twin Lakes, at an elevation of about 3048 m (10,000 ft.) (Scott 2004) (this is about where the red arrow points to Lake Creek on Fig. 8). Turner's specimens are undated, but they were most likely collected during the 1940s (that section of the old toll road was designated as part of US Hwy 82 in 1927; it was paved in 1967 (CDOT 2012)). Possibly in an attempt to further segregate pseudosamuelis, Brown et al. (1955b) limited this taxon to populations “generally above 11,500 feet,” ignoring the fact that the elevation of its type locality is 10,000 ft. Echoing Nabokov's analysis, Brown observed that specimens of melissa from the plains seem to “differ more from typical material caught at Fairplay than the Fairplay specimens differ from pseudosamuelis.”

Brown's notion of typical melissa was reinforced during the 1960s when he examined Mead's specimens at CMNH and discovered that they incorporated both high-elevation and low-elevation phenotypes. Edwards received and examined all of Mead's material from Colorado as it was collected, thus Brown considered Mead's entire catch to represent the type series (Brown 1970b). Agreeing with Nabokov (1949), Brown defined high-elevation populations as “true melissa” and low-elevation populations as the “better known atypical form.” That Edwards himself was aware of the variation in melissa is suggested by the figures in Mead ([1876]), which Edwards prepared under his personal direction. The ventral male is especially evocative of a high-elevation phenotype (Figs. 33, 34). In fact, Nabokov (1949) considered all the specimens figured by Mead ([1876]) to represent “typical” (i.e. high-elevation) melissa.

In selecting a lectotype of melissa, Brown worked with CMNH curator Harry K. Clench, a specialist of Lycaenidae. Brown (1970b) recalled that when reviewing lycaeinid material at CMNH, he and Clench “spent days upon days at the Carnegie Museum studying, discussing and arguing about this and that related to the selection of the Edwardsian specimens to be designated the type.” In November 1968 they selected one of three dated male specimens of melissa at CMNH which could be attributed to a high-elevation locality. Presumably because none of the available specimens at CMNH were from Park County, they chose one from another locality mentioned by Mead ([1876]): Twin Lakes (Lake County), which is the only high-elevation locality definitely represented among Mead's specimens in that collection (see Brown 1970b). Brown affixed a label to the specimen indicating that it was chosen as the lectotype by “F. M. Brown & H. K. Clench/XI 1968.”

The lectotype designation for melissa was based on a high-elevation concept as defined by Nabokov and supported by Clench, both of whom Brown recognized as authorities on this group of butterflies. This action, however, created a potential conflict with pseudosamuelis, which Brown (1970b) did not address. This is all the more surprising given that Brown believed that the lectotype was collected on La Plata Peak, a mountain over 4267 m (14,000 ft.) in elevation —well within the domain of pseudosamuelis as previously defined by Brown himself—and located only 4.8 km (3 mi) from the type locality of pseudosamuelis. Moreover, Nabokov (1949) had previously questioned if populations from Twin Lakes represented nominotypical melissa or another subspecies (i.e. pseudosamuelis). Undoubtedly because of the proximity of the type localities, Brown subsequently treated pseudosamuelis as a junior subjective synonym of melissa in Miller and Brown (1981). Perhaps bowing to criticism from those who preferred conventional usage, pseudosamuelis was again listed as a subspecies of melissa by Miller and Brown (1983).

Despite the interpretation of nominotypical melissa by Brown and Nabokov, Downey (1975) reintroduced the concept of this taxon as a “lowland, brightly colored” sagebrush-prairie inhabitant. He recognized the subspecies pseudosamuelis as “the extreme of an altitudinal cline.” Perhaps in protest, Downey made no mention of Brown's lectotype of melissa and instead reiterated the type locality of Nabokov (1949): “Park county, Colorado, between Fairplay and California Gulch.” Ironically, Downey's acceptance of a high-elevation type locality contradicted his premise and contributed to the confusion.

Lane and Weller (1994) mapped the distribution of pseudosamuelis to include a very small area surrounding its type locality in Lake County. Scott (2008a) collected additional examples of pseudosamuelis at its type locality and defined its range as “several of the cold valleys of the northern Sawatch Range.” Fisher (2009) observed that phenotypes resembling pseudosamuelis also occur at higher elevations in the southwestern mountains of Colorado. Scott (2006a) mentioned similar butterflies in the Bighorn Mountains of Wyoming, which he satirically referred to as “bighornimuelis.” Scott (2008a) later associated those populations with the taxon now recognized as Plebejus idas longinus (Nabokov). Although pseudosamuelis was reported from New Mexico by Ferris (1976) and Holland (1984), these records were not listed by Toliver et al. (2001). Further complicating matters is the extreme similarity between pseudosamuelis and the taxon currently recognized as Plebejus idas sublivens (Nabokov). Nabokov (1949, 1950) mentioned their resemblance and subsequent authors (e.g. Brown et al. 1955b, Scott 2006a, 2008a, Fisher 2009) postulated on their potential relationship. Gompert et al. (2010) found that idas and melissa have hybridized extensively in portions of the central Rocky Mountains, resulting in substantial admixture and nuclear introgression. Although authors continue to associate pseudosamuelis with melissa (Pelham 2008, 2014, Fisher 2009, Scott 2008a, 2014), it may be more closely allied with idas (A. D. Warren, pers. comm.). Scott (2008c) included all melissa subspecies within his own concept of an idas superspecies, which he sardonically termed a “stenchospecies.”

In an effort to fix the name melissa to a brightly colored lowland taxon, Scott (2006a, 2008c) asserted that Holland (1931) had designated a lectotype using a specimen from a low-elevation population. Holland referred to three figured specimens as “types:” a pair on Plate 31 and a female on Plate 66. Only the female on Plate 66 was identified as a type on the accompanying plate legend, which, Scott argued, demonstrates Holland's intent to select that specimen as the sole name-bearing type (i.e. lectotype) per ICZN (1999, Art. 74.5), thereby invalidating the lectotype designation by Brown (1970b). Scott (2006a) suggested a type locality of “Tinytown” (Jefferson County), which is about 2073 m (6800 ft.) in elevation. I disagree with this reasoning, as Holland did not unambiguously select the female to serve as the unique type as required by the Code (ICZN 1999, Art. 74.5), nor was there any visible intent to do so. What Scott interpreted as a conscious act by Holland was likely nothing more than an editorial artifact.

The revised edition of The Butterfly Book by Holland (1931) was based upon a first edition of 1898, which included 48 plates. Only two specimens figured in Holland (1898) were identified as types on the accompanying plate legends (Pl. 30, figs. 9, 33). For the revised edition of 1931, Holland added 29 additional plates, many of which included specimens that were identified as types on the plate legends. In only a few instances did he go back and add “type” captions to the original 48 plates. Simply stated, Holland (1931) did not identify the two specimens as types on the legend for Plate 31 (figs. 25, 26) because he had not done so in Holland (1898). Finally, there is no explicit statement on the legend for Plate 66 of Holland (1931) to suggest any intent to select that female as the sole name-bearing type. In the text of this book, Holland referred to three figured specimens as “types” (plural) without further restriction. The plate legend does not supersede the textual reference and instead merely denotes one of the three “types” mentioned in the text. I therefore agree with Pelham (2014) that the lectotype designation of Brown (1970b) is valid.