Matching Plant Images Internally

[Jorge_Stolfi]

Furthermore, the plants were also compared with classic medicinal plants, both today and in the past.

Yes, including by Petersen, in the ~1940s or earlier.

But consider f21r, for example.  Petersen had already "identified" it as "Polygonum" and you listed it above as You are not allowed to view links. Register or Login to view. (Pa) or knotgrass.  

At first sight the resemblance is remarkable and the identification seems to be a sure thing.  But the main similarity is due to f21r being drawn as viewed from the top, and that is how Pa is usually drawn, because of its habit of growing close to the ground.  Score one point for that.   Also both have long branching stems with small leaves all along them. Score another three points.  

But, if we look closely, we see several discrepancies:

• The leaves of Pa are generally pointy, while those of You are not allowed to view links. Register or Login to view. have mostly rounded ends.
  
• The leaves of Pa alternate along the stem, while those of You are not allowed to view links. Register or Login to view. are paired.
  
• The flowers of Pa grow singly or in small groups of 2-3 from the bases of the leaf stalk, forward of the leaf, while those of You are not allowed to view links. Register or Login to view. grow in dense bunches of a dozen or so flowers, directly out of the stem, right behind some leaf pairs.
  
• The flowers of Pa are elongated when still closed and with a trumpety corolla when open, while those of You are not allowed to view links. Register or Login to view. are spherical.
  
• The root of You are not allowed to view links. Register or Login to view. is a long and thick cylindrical "trunk", much wider than the stems, that branches out near the tip.  That does not seem to match the You are not allowed to view links. Register or Login to view.of Pa.
  

The distinction between alternate and paired leaves is generally a principal key for plant species identification.

I have seen one "herbal health" site where a picture labeled Pa has flowers in bunches that are more similar to those of f21r.  But it turns out that the site is wrong: their image is that of You are not allowed to view links. Register or Login to view. (Pm) or smartweed.  But even this is still not a perfect match, because the flowers  of Pm are elongated and the bunches apparently grow only at the end of long naked stalks. 

Sure, sure, those are drawings not photographs, and that can excuse some of the discrepancies above. (Although it is hard to believe that the Scribe would draw those flower bunches if he was looking at a real specimen.)  But they show how much discrepancy one must tolerate in order to get a match.  Even on the "best" cases...

All the best, --stolfi