ἔρημος or ἐρῆμος

The answer can be found in Blass-Debrunner-Funk's grammar.  I do not own the Logos version, but in my print version (1961) it can be found on page 9 (section 13).  This article states that the form ἐρῆμος is "distinctly ancient and foreign to Koine".  I guess it is sort of like the difference between "wherefore" and "why".

It's exactly as I said, there's something not quite right with the search engine...

Personally, I hope Logos will fix the parts that need fixing. It should be a priority.

(UPDATING earlier post with new possibilities. I will also cross-post with the earlier thread.)

I noted in the other thread thatI discovered that separate searches had to be done for λεγω and ειπον and for οραω and ειδον. A bit annoying, especially if you don't know that these distinctions exist, but understandable in light of how the language developed and how BDAG handles such matters.

So... what I've become accustomed to doing is not simply trusting the right-click on a word to choose to search on lemma. Instead, if I am wanting to be comprehensive, I have some options:

1. Choose to search on root rather than lemma. The downside to this is that it may give me a lot of hits I don't want. Additionally, I can't see that any of the LXX texts in Logos have tagged roots, so you can't use it for the LXX.
2. Instead of right-click, simply use the search box dialogue. Start typing, and note the various forms that show up. In the example which initiated this thread, both of the ways of accenting ερημος show up in the dropdown box, so I know to search for: lemma:ἐρῆμος OR lemma:ἔρημος
   This solution works well except for cases like λεγω or οραω noted above where a different stem is used for different tenses.
3. Search on the NT in the Original Greek Byzantine Textform 2005 (Byz) or on Westcott-Hort Greek NT 1881 (WH1881MR; not the Logos WH). These are coded differently and don't make the same lemma or accent distinctions.
4. Create a collection of Greek NT versions. You can right-click and conduct a lemma search. Then, in the search results, change the specific version to that collection of Greek NT versions you created. Then use the grid view to spot right away where there might be differences in parsing or textual tradition. (Cf. below)
5. I use BibleWorks which doesn't make these distinctions. (And remember that can be both good or bad depending on what you exactly want.)

I want to highlight how useful and valuable option #4 is. Here is what a search on λεγω returns:

As you can easily see, something is going on with the Byz and WH1881MR that immediately highlights that there are perhaps coding issues to which I should attend.

Note that the Grid view is not available in a Morph search, but you can use the Analysis view or Aligned view and with just a bit of comparing see rather quickly if there are any issues.

This method is also a great way of seeing if there are text critical issues. For example, a lemma search for σπλαγχνιζομαι will return:

What should catch your attention right away is that missing entry in Mark 1.41 of the SBLGNT, a somewhat well-known and notable textual variant.

The only thing that complicates matters here for me is that I am not entirely confident about constructing search strings from scratch. (Cf. in the graphics above the search string that was used.) I find it easier to right-click on a word in a text to start my search, then in the search results window, change the parameters I want: what passages, in what versions/collection, Bible or morph search, etc.