After trying to collect some data about the loco construction numbers following the old ALCO amalgamation in the USA in 1901, I have come to the conclusion that the "official" list may not be correct at all! I e, that the records of the smaller shops were not always adhered to, when it came to compiling a complete and all-inclusive list of engines built. The main headache here seems to be the Richmond works, which employed very high c/n´s despite their not building very many locomotives (according to certain sources)!
The reason I noticed this is, that RLW (reportedly) assigned nos. 2999 to 3001 to three engines delivered to a private RR in Sweden in 1900. (These in fact survived well into the 1950s, when interest in these matters had begun to grow - I therefore assume the numbers to be correct. A batch of other engines supplied by RLW to our State Rys. the year before also had c/n´s in the 2800 series... That we know for a fact.)
Thus my question is: Did the ALCO constr. numbering really take this into account? RLW are supposed to have built just some 1,000 locomotives!!!