Help with formula

[null null]

I have a list of cities (column B) and the number of companies (column G) that based in these cities. The column B has duplicates for example:
Canterbury
Canterbury , Kent
Canterbury, Kent

Column D (that represent a number of these companies) linked to the column B:

3
1
39

To make sure that the city located at the specific area, I have added it's latitude and longitude columns (C and D). Having in mind that the latitude and longitude data is identical I'm 100% sure that "Canterbury", "Canterbury , Kent" and "Canterbury, Kent" is the same place. How I can get the total of companies based in Canterbury (43) using columns B, C, D and G? What formula do I have to use? Please direct.

[#2]

Could be

=SUMIFS(G:G,C:C,INDEX(C:C,MATCH("Canterbury",B:B,0)),D:D,INDEX(D:D,MATCH("Canterbury",B:B,0)))

[#3]

Thank you Sergei, but this won't work. I have attached the spreadsheet for better understanding. Any help would be greatly appreciated.

[#4]

I think you shall identify you addresses first by one common name for each address. You may add helper column like

=INDEX([Address], MATCH(1,INDEX(([@Easting]=[Easting])*([@Northing]=[Northing]),0,1),0))

which gives one name (first match) for the places with same coordinates. 

After that Pivot Table gives you the sum of companies for each place. Please see attached.

[#5]

Sergei, words are powerless to express my gratitude. THANK YOU SOO MUCH !!!

[#6]

I had the same job to do for a different goal, so I create a list of things to do to accomplish that mission.

I attached the file; this way you don't have to rename all 2000 cities, but it checks it with all the coordinates (which I assume are correct).

Columns in green are the columns you have normally.

Red column is the important one, it merges all the coordinates.

Blue columns are a copy of the green ones, but all datas in that columns are going to be displayed by a formula.

Column J checks if the coordinates are a duplicate, if that so, it leave the cell blank.

All the rest of the columns give something different than an empty cell only if the correspondant J cell is not empty (i.e.: if the row is not a duplicate).

Column P sum all the # of companies of the same coordinates in the first appear of the coordinates.

When you copypaste these formulas on all 2000 rows, you MUST copy columns J to P and paste them as valor (special paste), removing formulas. Then you can delete columns A to I and have a "new table" with some blank rows and the right sum of all the duplicates.

Then sort it or filter it to remove blank rows and the mission is accomplished!