Here we present the digitized timeseries of hourly observations of the geomagnetic field declination (D) in Irkutsk in 1890 from unpublished handwritten monthly tables. Observations of declination variations were made with the Edelmann and Krause monofilar magnetometers. These data from the Irkutsk Observatory (IRT code) were not initially included in the published annual reports. All date/time marks are converted to UT. Declination is given in minutes of arc. The original digital images of the tables are included as well.For reference, we also provide the digitized table as it appears in the original handwritten record (D-1890-source.dat). The data were digitized manually; first column of the table contains date, next 24 columns are hourly values in minutes of arc, and the last column is true daily average (TDA). In fact, all hourly quantities in the table are not true declination values, but deviations from daily averages. To obtain a declination value, one should take an hourly value from table, add a corresponding daily average (TDA) and a constant offset term (2 degrees, or 120 minutes). Another complication comes from the fact that westerly variation, unlike today, was considered positive. Thus, for 1890-01-01T01:00 true (easterly) declination value should be -(-0.2+14.08+120)=-133.88 minutes of arc, and so on. You may notice that hourly values start at 1AM and finish at midnight, so midnight refers to the previous day. This was a common convention at that time. All time marks in the source table are local (geographical location of the Observatory was 52°16'19N, 104°18'27E); time shift is 7 hours. Keep in mind that an Orthodox calendar was in use in Russian Empire at those times, so, date shift for the year 1890 was 12 days.

Additional metadata file descriptions: img.zip contains scanned digital images of original historical records.D-1890-source.dat is a digitized table of the original handwritten record.source2ts.jl is a Julia script for conversion of the source table to timeseries.