MongoDB GeoSpatial Query on subdocument C#

Question

I've successfully got

 var locations = MongoCollection.Find(Query.Within<GeoJson2DGeographicCoordinates>("Position", poly))

to return the correct results when my document has this structure:

class Location
{
    [BsonId]
    public int Id { get; set; }
    public string Address { get; set; }
    public GeoJsonPoint<GeoJson2DGeographicCoordinates> Position { get; set; }
}

and this index:

MongoCollection.CreateIndex(IndexKeys.GeoSpatialSpherical("Position"));

However, the minute I change my document to contain a sub document with an array of Foos like this:

 class Location
    {
        [BsonId]
        public int Id { get; set; }
        public List<Foo> Locations {get;set;}
    }

class Foo
       {
            public int Id { get; set; }
            public string Address { get; set; }
            public GeoJsonPoint<GeoJson2DGeographicCoordinates> Position { get; set; }
        }  

and I change the index and query to use the sub document fields like so:

MongoCollection.CreateIndex(IndexKeys.GeoSpatialSpherical("Locations.Position"));    
var locations = MongoCollection.Find(Query.Within<GeoJson2DGeographicCoordinates>("Locations.Position", poly)).AsQueryable().ToList();

It stops working as expected. In the above scenario it returns back a list of Location with 1 item, that item is all the Foos.

Updated To elaborate, if i have 400 Foo objects, then after i run the query on the sub-document it returns me a List where Locations is a List with 400 Foos. So it's not successfully checking if the Foos are in the poly.

From my immediate window:

locations
Count = 1
    [0]: {MyGeo.GeoTest.Location}
locations[0].Locations
Count = 408

What I expect is for List Locations to be filtered based on the poly.

Can anyone tell me where i'm going wrong? I've searched online and there doesn't seem to be a lot of examples with this structure but the documentation says you can have the index the way i have it.

You may also model the location data as a field inside of a sub-document. In this case, the document would contain a field (e.g. addresses) that holds an array of documents where each document has a field (e.g. loc:) that holds location coordinates. For example:

{ _id : ObjectId(...),
  name : "...",
  addresses : [ {
                 context : "home" ,
                 loc : [ 55.5, 42.3 ]
                } ,
                {
                 context : "home",
                 loc : [ -74 , 44.74 ]
                }
              ]
}

You could then create the geospatial index on the addresses.loc field as in the following example:

db.records.ensureIndex( { "addresses.loc": "2d" } )

My query must be wrong because i must not be understanding how the sub documents work correctly.

Any ideas?

Thanks.