I am out at the Herbarium with the human female today. She is, as she likes to call it, “hunting bogeys“—finding and flagging errors in the herbarium database, finding the actual specimens in the collection, and correcting the errors.
The database, some 192,000 records in all, was cobbled together from multiple smaller source files. Over the years, she has put in countless hours, correcting collector names; fixing plant family designations; correcting misspellings in county, location, and habitat information; splitting out elevation, plant description, longitude and latitude, and additional notes into their own fields, and the like. Meanwhile, I am hard at work corrupting scientific names, blurring old handwriting, and whispering strange and capricious assumptions into the ears of the student workers preparing the next batch of records slated to be added.
Probably my best trick was teaching the student workers to just record the last two digits of the collection year. When the output from the old software was brought into Excel, I nudged Excel to randomly assign a century to go with those two digits, so that there was produced a mangled melange of 19th-century specimens sporting collection years such as 1988, 2000, 2043, 2077, and 2068. There are also 21st century specimens listed as having been collected in 1902, 1907, and 1910, etc. I am really proud of this particular bit of mischief because I did it so that there is no pattern to the errors such that a blanket fix may be imposed. I mean, once a collector is pegged to the proper century, all of his record years can be adjusted without looking each one up, but I made a LOT of one-offs for collectors without a lot of specimens that must still be checked one … by … one.
So that’s what she is working on today, hunting up those bad-year specimens and correcting the records. While she’s at it, she’s fixing other problems. She’s taken to marking corrections in green, just so she can look at the stupid database and feel like she’s making progress. (Oh, the fallible mortal need for constant affirmation!)
For example, here’s a sedge-y thing that some enterprising student worker, at my urging, databased as having been collected by J. K. Wipff. As you can see, he was merely an annotator, some thirty-four years after the fact.
There. Now it’s correct in the database. W. G. Dore. Good old W. G. Bet he used to like to tell people it stood for, “Well, golly.” Or perhaps, since he worked for the Department of Agriculture, perhaps it was short for “Weed guru.” It’s so amusing to speculate.
Great Frigga’s Hairpins! Look at this one.
Ehehehehe! I mean, tsk, tsk tsk!
Periploca is not in the Dioscoreaceaea. It’s not even a monocot! I’ll let her correct it to Asclepiadaceae while I go tug something else out of alignment.
Oh, now here’s a nice plant. Yellow flowers, spiffy legume fruit. Interesting provenance.
It belongs to the genus Senna. Senna is one of the smaller genera separated from the huge, catch-all genus Cassia. Cassia was just too bloated—it contained several different entities that really are best treated as separate entities. I saw to it that this one specimen was treated as all three.
It was collected fairly recently, so the original label clearly says “Senna,” which is correct.
However, it was databased as Cassia and filed in Chamaecrista, another segregate genus.
In a folder with a misspelled country:
The student workers didn’t really need much help. They did a lot on their own. Here’s a good one. The label says “New Hanover Co., ” which is in North Carolina. I had the student worker put it in as “New Haven, Connecticut.” Here’s a label in German, which was databased as collected in Germany, even though it’s from the U.S. And here’s another, databased and filed as African, though it’s from the canton of Valois, Switzerland.
And this little gem was databased as U. S., when it is from Natal, which is now part of a realm called “South Africa.”
Keep up the fine work, student workers! All we have to do is work slightly faster than the human female to assure that this project is never,