Why~400 FPS difference between these two methods of building a vertex array

[IGD]

When I build my vertex array using this code, I get like 2-350 FPS:

                sf::Vertex v0(sf::Vector2f(position.x * size.x, position.y * size.y), sf::Vector2f((tu * size.x)+5, (tv * size.y)+5));
                sf::Vertex v1(sf::Vector2f((position.x + 1) * size.x, position.y * size.y), sf::Vector2f(((tu + 1) * size.x)-5, (tv * size.y)+5));
                sf::Vertex v2(sf::Vector2f(position.x * size.x, (position.y + 1) * size.y), sf::Vector2f((tu * size.x)+5, ((tv + 1) * size.y)-5));
                sf::Vertex v3(sf::Vector2f(position.x * size.x, (position.y + 1) * size.y), sf::Vector2f((tu * size.x)+5, ((tv + 1) * size.y)-5));
                sf::Vertex v4(sf::Vector2f((position.x + 1) * size.x, position.y * size.y), sf::Vector2f(((tu + 1) * size.x)-5, (tv * size.y)+5));
                sf::Vertex v5(sf::Vector2f((position.x + 1) * size.x, (position.y + 1) * size.y), sf::Vector2f(((tu + 1) * size.x)-5, ((tv + 1) * size.y)-5));


                m_vertices.append(v0);
                m_vertices.append(v1);
                m_vertices.append(v2);
                m_vertices.append(v3);
                m_vertices.append(v4);
                m_vertices.append(v5);
 

BUT when I build that same vertex array with this code, I get 4-800 FPS:

sf::Vertex* triangles = &m_vertices[(i + j * width) * 6];

                // define the 6 corners of the two triangles
                triangles[0].position = sf::Vector2f(i * tileSize.x, j * tileSize.y);
                triangles[1].position = sf::Vector2f((i + 1) * tileSize.x, j * tileSize.y);
                triangles[2].position = sf::Vector2f(i * tileSize.x, (j + 1) * tileSize.y);
                triangles[3].position = sf::Vector2f(i * tileSize.x, (j + 1) * tileSize.y);
                triangles[4].position = sf::Vector2f((i + 1) * tileSize.x, j * tileSize.y);
                triangles[5].position = sf::Vector2f((i + 1) * tileSize.x, (j + 1) * tileSize.y);

                // define the 6 matching texture coordinates
                triangles[0].texCoords = sf::Vector2f(tu * tileSize.x+5, tv * tileSize.y+5);
                triangles[1].texCoords = sf::Vector2f((tu + 1) * tileSize.x-5, tv * tileSize.y+5);
                triangles[2].texCoords = sf::Vector2f(tu * tileSize.x+5, (tv + 1) * tileSize.y-5);
                triangles[3].texCoords = sf::Vector2f(tu * tileSize.x+5, (tv + 1) * tileSize.y-5);
                triangles[4].texCoords = sf::Vector2f((tu + 1) * tileSize.x-5, tv * tileSize.y+5);
                triangles[5].texCoords = sf::Vector2f((tu + 1) * tileSize.x-5, (tv + 1) * tileSize.y-5);
 

I would understand the FPS difference if the function was called each loop, but it is only called once when initializing. Why in the world is the FPS so different???

[IGD]

So I continued digging, and I found this:
https://en.sfml-dev.org/forums/index.php?topic=22780.0

Upon looking at my code deeper, I noticed I was in fact resizing the vertex array before utilizing append, which, as was the case with the previous poster, led to double the vertices.